Artificial Intelligence Nanodegree¶
Convolutional Neural Networks¶
Project: Write an Algorithm for a Dog Identification App¶
In this notebook, some template code has already been provided for you, and you will need to implement additional functionality to successfully complete this project. You will not need to modify the included code beyond what is requested. Sections that begin with '(IMPLEMENTATION)' in the header indicate that the following block of code will require additional functionality which you must provide. Instructions will be provided for each section, and the specifics of the implementation are marked in the code block with a 'TODO' statement. Please be sure to read the instructions carefully!
Note: Once you have completed all of the code implementations, you need to finalize your work by exporting the iPython Notebook as an HTML document. Before exporting the notebook to html, all of the code cells need to have been run so that reviewers can see the final implementation and output. You can then export the notebook by using the menu above and navigating to \n", "File -> Download as -> HTML (.html). Include the finished document along with this notebook as your submission.
In addition to implementing code, there will be questions that you must answer which relate to the project and your implementation. Each section where you will answer a question is preceded by a 'Question X' header. Carefully read each question and provide thorough answers in the following text boxes that begin with 'Answer:'. Your project submission will be evaluated based on your answers to each of the questions and the implementation you provide.
Note: Code and Markdown cells can be executed using the Shift + Enter keyboard shortcut. Markdown cells can be edited by double-clicking the cell to enter edit mode.
The rubric contains optional "Stand Out Suggestions" for enhancing the project beyond the minimum requirements. If you decide to pursue the "Stand Out Suggestions", you should include the code in this IPython notebook.
Why We're Here¶
In this notebook, you will make the first steps towards developing an algorithm that could be used as part of a mobile or web app. At the end of this project, your code will accept any user-supplied image as input. If a dog is detected in the image, it will provide an estimate of the dog's breed. If a human is detected, it will provide an estimate of the dog breed that is most resembling. The image below displays potential sample output of your finished project (... but we expect that each student's algorithm will behave differently!).
In this real-world setting, you will need to piece together a series of models to perform different tasks; for instance, the algorithm that detects humans in an image will be different from the CNN that infers dog breed. There are many points of possible failure, and no perfect algorithm exists. Your imperfect solution will nonetheless create a fun user experience!
The Road Ahead¶
We break the notebook into separate steps. Feel free to use the links below to navigate the notebook.
- Step 0: Import Datasets
- Step 1: Detect Humans
- Step 2: Detect Dogs
- Step 3: Create a CNN to Classify Dog Breeds (from Scratch)
- Step 4: Use a CNN to Classify Dog Breeds (using Transfer Learning)
- Step 5: Create a CNN to Classify Dog Breeds (using Transfer Learning)
- Step 6: Write your Algorithm
- Step 7: Test Your Algorithm
Step 0: Import Datasets¶
Import Dog Dataset¶
In the code cell below, we import a dataset of dog images. We populate a few variables through the use of the load_files function from the scikit-learn library:
train_files,valid_files,test_files- numpy arrays containing file paths to imagestrain_targets,valid_targets,test_targets- numpy arrays containing onehot-encoded classification labelsdog_names- list of string-valued dog breed names for translating labels
In [1]:
from sklearn.datasets import load_files
from keras.utils import np_utils
import numpy as np
from glob import glob
# define function to load train, test, and validation datasets
def load_dataset(path):
data = load_files(path)
dog_files = np.array(data['filenames'])
dog_targets = np_utils.to_categorical(np.array(data['target']), 133)
return dog_files, dog_targets
# load train, test, and validation datasets
train_files, train_targets = load_dataset('dogImages/train')
valid_files, valid_targets = load_dataset('dogImages/valid')
test_files, test_targets = load_dataset('dogImages/test')
# load list of dog names
dog_names = [item[20:-1] for item in sorted(glob("dogImages/train/*/"))]
# print statistics about the dataset
print('There are %d total dog categories.' % len(dog_names))
print('There are %s total dog images.\n' % len(np.hstack([train_files, valid_files, test_files])))
print('There are %d training dog images.' % len(train_files))
print('There are %d validation dog images.' % len(valid_files))
print('There are %d test dog images.'% len(test_files))
Using TensorFlow backend.
There are 133 total dog categories. There are 8351 total dog images. There are 6680 training dog images. There are 835 validation dog images. There are 836 test dog images.
Import Human Dataset¶
In the code cell below, we import a dataset of human images, where the file paths are stored in the numpy array human_files.
In [2]:
import random
random.seed(8675309)
# load filenames in shuffled human dataset
human_files = np.array(glob("lfw/*/*"))
random.shuffle(human_files)
# print statistics about the dataset
print('There are %d total human images.' % len(human_files))
There are 13233 total human images.
Step 1: Detect Humans¶
We use OpenCV's implementation of Haar feature-based cascade classifiers to detect human faces in images. OpenCV provides many pre-trained face detectors, stored as XML files on github. We have downloaded one of these detectors and stored it in the haarcascades directory.
In the next code cell, we demonstrate how to use this detector to find human faces in a sample image.
In [3]:
import cv2
import matplotlib.pyplot as plt
%matplotlib inline
# extract pre-trained face detector
face_cascade = cv2.CascadeClassifier('haarcascades/haarcascade_frontalface_alt.xml')
# load color (BGR) image
img = cv2.imread(human_files[3])
# convert BGR image to grayscale
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
# find faces in image
faces = face_cascade.detectMultiScale(gray)
# print number of faces detected in the image
print('Number of faces detected:', len(faces))
# get bounding box for each detected face
for (x,y,w,h) in faces:
# add bounding box to color image
cv2.rectangle(img,(x,y),(x+w,y+h),(255,0,0),2)
# convert BGR image to RGB for plotting
cv_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
# display the image, along with bounding box
plt.imshow(cv_rgb)
plt.show()
Number of faces detected: 1
Before using any of the face detectors, it is standard procedure to convert the images to grayscale. The detectMultiScale function executes the classifier stored in face_cascade and takes the grayscale image as a parameter.
In the above code, faces is a numpy array of detected faces, where each row corresponds to a detected face. Each detected face is a 1D array with four entries that specifies the bounding box of the detected face. The first two entries in the array (extracted in the above code as x and y) specify the horizontal and vertical positions of the top left corner of the bounding box. The last two entries in the array (extracted here as w and h) specify the width and height of the box.
Write a Human Face Detector¶
We can use this procedure to write a function that returns True if a human face is detected in an image and False otherwise. This function, aptly named face_detector, takes a string-valued file path to an image as input and appears in the code block below.
In [4]:
# returns "True" if face is detected in image stored at img_path
def face_detector(img_path):
img = cv2.imread(img_path)
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(gray)
return len(faces) > 0
(IMPLEMENTATION) Assess the Human Face Detector¶
Question 1: Use the code cell below to test the performance of the face_detector function.
- What percentage of the first 100 images in
human_fileshave a detected human face? - What percentage of the first 100 images in
dog_fileshave a detected human face?
Ideally, we would like 100% of human images with a detected face and 0% of dog images with a detected face. You will see that our algorithm falls short of this goal, but still gives acceptable performance. We extract the file paths for the first 100 images from each of the datasets and store them in the numpy arrays human_files_short and dog_files_short.
Answer:
In [5]:
import cv2
human_files_short = human_files[:100]
dog_files_short = train_files[:100]
# Do NOT modify the code above this line.
## TODO: Test the performance of the face_detector algorithm
## on the images in human_files_short and dog_files_short.
count=0
for ii in range (len(human_files_short)):
face_h=face_detector(human_files_short[ii])
if (face_h >0):
count=count+1
print(str(count), "% of the humans were detected as humans!")
#count=0
#for ii in range (len(dog_files_short)):
# face_d=face_detector(dog_files_short[ii])
# if (face_d ==1):
# count=count+1
#print(str(count), "% of the dogs were detected as humans!")
99 % of the humans were detected as humans!
Question 2: This algorithmic choice necessitates that we communicate to the user that we accept human images only when they provide a clear view of a face (otherwise, we risk having unneccessarily frustrated users!). In your opinion, is this a reasonable expectation to pose on the user? If not, can you think of a way to detect humans in images that does not necessitate an image with a clearly presented face?
Answer:
We suggest the face detector from OpenCV as a potential way to detect human images in your algorithm, but you are free to explore other approaches, especially approaches that make use of deep learning :). Please use the code cell below to design and test your own face detection algorithm. If you decide to pursue this optional task, report performance on each of the datasets.
In [6]:
## (Optional) TODO: Report the performance of another
## face detection algorithm on the LFW dataset
### Feel free to use as many code cells as needed.
#import OpenCV module
import cv2
#import os module for reading training data directories and paths
import os
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
#import numpy to convert python lists to numpy arrays as
#it is needed by OpenCV face recognizers
#function to detect face using OpenCV
face_cascade = cv2.CascadeClassifier('haarcascades/haarcascade_frontalface_alt.xml')
human_files_short = human_files[:100]
for i in range(len(human_files_short)):
img = cv2.imread(human_files_short[i])
#convert the test image to gray image as opencv face detector expects gray images
gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
#load OpenCV face detector, using LBP which is fast
#there is also a more accurate but slow Haar classifier
#let's detect multiscale (some images may be closer to camera than others) images
#result is a list of faces
faces = face_cascade.detectMultiScale(gray, scaleFactor=1.2, minNeighbors=5)
#under the assumption that there will be only one face,
#extract the face area
for (x,y,w,h) in faces:
cv2.rectangle(img,(x,y),(x+w,y+h),(255,0,0),2)
cv_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
plt.imshow(cv_rgb)
plt.show()
if (len(faces) == 0):
print('No faces detected')
Step 2: Detect Dogs¶
In this section, we use a pre-trained ResNet-50 model to detect dogs in images. Our first line of code downloads the ResNet-50 model, along with weights that have been trained on ImageNet, a very large, very popular dataset used for image classification and other vision tasks. ImageNet contains over 10 million URLs, each linking to an image containing an object from one of 1000 categories. Given an image, this pre-trained ResNet-50 model returns a prediction (derived from the available categories in ImageNet) for the object that is contained in the image.
In [7]:
from keras.applications.resnet50 import ResNet50
# define ResNet50 model
ResNet50_model = ResNet50(weights='imagenet')
Pre-process the Data¶
When using TensorFlow as backend, Keras CNNs require a 4D array (which we'll also refer to as a 4D tensor) as input, with shape
$$ (\text{nb_samples}, \text{rows}, \text{columns}, \text{channels}), $$
where nb_samples corresponds to the total number of images (or samples), and rows, columns, and channels correspond to the number of rows, columns, and channels for each image, respectively.
The path_to_tensor function below takes a string-valued file path to a color image as input and returns a 4D tensor suitable for supplying to a Keras CNN. The function first loads the image and resizes it to a square image that is $224 \times 224$ pixels. Next, the image is converted to an array, which is then resized to a 4D tensor. In this case, since we are working with color images, each image has three channels. Likewise, since we are processing a single image (or sample), the returned tensor will always have shape
$$ (1, 224, 224, 3). $$
The paths_to_tensor function takes a numpy array of string-valued image paths as input and returns a 4D tensor with shape
$$ (\text{nb_samples}, 224, 224, 3). $$
Here, nb_samples is the number of samples, or number of images, in the supplied array of image paths. It is best to think of nb_samples as the number of 3D tensors (where each 3D tensor corresponds to a different image) in your dataset!
In [8]:
from keras.preprocessing import image
from tqdm import tqdm
def path_to_tensor(img_path):
# loads RGB image as PIL.Image.Image type
img = image.load_img(img_path, target_size=(224, 224))
# convert PIL.Image.Image type to 3D tensor with shape (224, 224, 3)
x = image.img_to_array(img)
# convert 3D tensor to 4D tensor with shape (1, 224, 224, 3) and return 4D tensor
return np.expand_dims(x, axis=0)
def paths_to_tensor(img_paths):
list_of_tensors = [path_to_tensor(img_path) for img_path in tqdm(img_paths)]
return np.vstack(list_of_tensors)
Making Predictions with ResNet-50¶
Getting the 4D tensor ready for ResNet-50, and for any other pre-trained model in Keras, requires some additional processing. First, the RGB image is converted to BGR by reordering the channels. All pre-trained models have the additional normalization step that the mean pixel (expressed in RGB as $[103.939, 116.779, 123.68]$ and calculated from all pixels in all images in ImageNet) must be subtracted from every pixel in each image. This is implemented in the imported function preprocess_input. If you're curious, you can check the code for preprocess_input here.
Now that we have a way to format our image for supplying to ResNet-50, we are now ready to use the model to extract the predictions. This is accomplished with the predict method, which returns an array whose $i$-th entry is the model's predicted probability that the image belongs to the $i$-th ImageNet category. This is implemented in the ResNet50_predict_labels function below.
By taking the argmax of the predicted probability vector, we obtain an integer corresponding to the model's predicted object class, which we can identify with an object category through the use of this dictionary.
In [9]:
from keras.applications.resnet50 import preprocess_input, decode_predictions
def ResNet50_predict_labels(img_path):
# returns prediction vector for image located at img_path
img = preprocess_input(path_to_tensor(img_path))
return np.argmax(ResNet50_model.predict(img))
In [10]:
### returns "True" if a dog is detected in the image stored at img_path
def dog_detector(img_path):
prediction = ResNet50_predict_labels(img_path)
return ((prediction <= 268) & (prediction >= 151))
Write a Dog Detector¶
While looking at the dictionary, you will notice that the categories corresponding to dogs appear in an uninterrupted sequence and correspond to dictionary keys 151-268, inclusive, to include all categories from 'Chihuahua' to 'Mexican hairless'. Thus, in order to check to see if an image is predicted to contain a dog by the pre-trained ResNet-50 model, we need only check if the ResNet50_predict_labels function above returns a value between 151 and 268 (inclusive).
We use these ideas to complete the dog_detector function below, which returns True if a dog is detected in an image (and False if not).
(IMPLEMENTATION) Assess the Dog Detector¶
Question 3: Use the code cell below to test the performance of your dog_detector function.
- What percentage of the images in
human_files_shorthave a detected dog? - What percentage of the images in
dog_files_shorthave a detected dog?
Answer:
In [11]:
### TODO: Test the performance of the dog_detector function
### on the images in human_files_short and dog_files_short.
import cv2
human_files_short = human_files[:100]
dog_files_short = train_files[:100]
count=0
for ii in range (len(human_files_short)):
dog_h=dog_detector(human_files_short[ii])
if (dog_h >0):
count=count+1
print(str(count) + "% of the humans were detected as dogs")
count=0
for ii in range (len(dog_files_short)):
dog_d=dog_detector(dog_files_short[ii])
if (dog_d >0):
count=count+1
print(str(count) + "% of the dogs were detected as dogs")
1% of the humans were detected as dogs 100% of the dogs were detected as dogs
Step 3: Create a CNN to Classify Dog Breeds (from Scratch)¶
Now that we have functions for detecting humans and dogs in images, we need a way to predict breed from images. In this step, you will create a CNN that classifies dog breeds. You must create your CNN from scratch (so, you can't use transfer learning yet!), and you must attain a test accuracy of at least 1%. In Step 5 of this notebook, you will have the opportunity to use transfer learning to create a CNN that attains greatly improved accuracy.
Be careful with adding too many trainable layers! More parameters means longer training, which means you are more likely to need a GPU to accelerate the training process. Thankfully, Keras provides a handy estimate of the time that each epoch is likely to take; you can extrapolate this estimate to figure out how long it will take for your algorithm to train.
We mention that the task of assigning breed to dogs from images is considered exceptionally challenging. To see why, consider that even a human would have great difficulty in distinguishing between a Brittany and a Welsh Springer Spaniel.
| Brittany | Welsh Springer Spaniel |
|---|---|
 |
 |
It is not difficult to find other dog breed pairs with minimal inter-class variation (for instance, Curly-Coated Retrievers and American Water Spaniels).
| Curly-Coated Retriever | American Water Spaniel |
|---|---|
 |
 |
Likewise, recall that labradors come in yellow, chocolate, and black. Your vision-based algorithm will have to conquer this high intra-class variation to determine how to classify all of these different shades as the same breed.
| Yellow Labrador | Chocolate Labrador | Black Labrador |
|---|---|---|
 |
 |
 |
We also mention that random chance presents an exceptionally low bar: setting aside the fact that the classes are slightly imabalanced, a random guess will provide a correct answer roughly 1 in 133 times, which corresponds to an accuracy of less than 1%.
Remember that the practice is far ahead of the theory in deep learning. Experiment with many different architectures, and trust your intuition. And, of course, have fun!
Pre-process the Data¶
We rescale the images by dividing every pixel in every image by 255.
In [12]:
from PIL import ImageFile
ImageFile.LOAD_TRUNCATED_IMAGES = True
# pre-process the data for Keras
train_tensors = paths_to_tensor(train_files).astype('float32')/255
valid_tensors = paths_to_tensor(valid_files).astype('float32')/255
test_tensors = paths_to_tensor(test_files).astype('float32')/255
100%|██████████████████████████████████████████████████████████████████████████████| 6680/6680 [01:50<00:00, 60.23it/s] 100%|████████████████████████████████████████████████████████████████████████████████| 835/835 [00:16<00:00, 50.33it/s] 100%|████████████████████████████████████████████████████████████████████████████████| 836/836 [00:15<00:00, 55.47it/s]
(IMPLEMENTATION) Model Architecture¶
Create a CNN to classify dog breed. At the end of your code cell block, summarize the layers of your model by executing the line:
model.summary()
We have imported some Python modules to get you started, but feel free to import as many modules as you need. If you end up getting stuck, here's a hint that specifies a model that trains relatively fast on CPU and attains >1% test accuracy in 5 epochs:
Question 4: Outline the steps you took to get to your final CNN architecture and your reasoning at each step. If you chose to use the hinted architecture above, describe why you think that CNN architecture should work well for the image classification task.
Answer:
In [13]:
from keras.layers import Conv2D, MaxPooling2D, GlobalAveragePooling2D
from keras.layers import Dropout, Flatten, Dense
from keras.models import Sequential
model = Sequential()
model.add(Conv2D(filters=16,kernel_size=2,padding='same',activation='relu', input_shape=(224,224,3)))
model.add(MaxPooling2D(pool_size=2))
model.add(Conv2D(filters=32,kernel_size=2,padding='same',activation='relu'))
model.add(MaxPooling2D(pool_size=2))
model.add(Conv2D(filters=64,kernel_size=2,padding='same',activation='relu'))
model.add(MaxPooling2D(pool_size=2))
model.add(Dropout(0.3))
model.add(Flatten())
model.add(Dense(500,activation='relu'))
model.add(Dropout(0.4))
model.add(Dense(133, activation='softmax'))
model.summary()
_________________________________________________________________ Layer (type) Output Shape Param # ================================================================= conv2d_1 (Conv2D) (None, 224, 224, 16) 208 _________________________________________________________________ max_pooling2d_2 (MaxPooling2 (None, 112, 112, 16) 0 _________________________________________________________________ conv2d_2 (Conv2D) (None, 112, 112, 32) 2080 _________________________________________________________________ max_pooling2d_3 (MaxPooling2 (None, 56, 56, 32) 0 _________________________________________________________________ conv2d_3 (Conv2D) (None, 56, 56, 64) 8256 _________________________________________________________________ max_pooling2d_4 (MaxPooling2 (None, 28, 28, 64) 0 _________________________________________________________________ dropout_1 (Dropout) (None, 28, 28, 64) 0 _________________________________________________________________ flatten_2 (Flatten) (None, 50176) 0 _________________________________________________________________ dense_1 (Dense) (None, 500) 25088500 _________________________________________________________________ dropout_2 (Dropout) (None, 500) 0 _________________________________________________________________ dense_2 (Dense) (None, 133) 66633 ================================================================= Total params: 25,165,677.0 Trainable params: 25,165,677.0 Non-trainable params: 0.0 _________________________________________________________________
Compile the Model¶
In [14]:
model.compile(optimizer='rmsprop', loss='categorical_crossentropy', metrics=['accuracy'])
(IMPLEMENTATION) Train the Model¶
Train your model in the code cell below. Use model checkpointing to save the model that attains the best validation loss.
You are welcome to augment the training data, but this is not a requirement.
Load the Model with the Best Validation Loss¶
In [15]:
from keras.callbacks import ModelCheckpoint
### TODO: specify the number of epochs that you would like to use to train the model.
epochs = 5
### Do NOT modify the code below this line.
checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.from_scratch.hdf5',
verbose=1, save_best_only=True)
model.fit(train_tensors, train_targets,
validation_data=(valid_tensors, valid_targets),
epochs=epochs, batch_size=20, callbacks=[checkpointer], verbose=1)
Train on 6680 samples, validate on 835 samples Epoch 1/5 4040/6680 [=================>............] - ETA: 2530s - loss: 4.9600 - acc: 0.0000e+ - ETA: 2524s - loss: 9.8054 - acc: 0.0250 - ETA: 3191s - loss: 11.2842 - acc: 0.033 - ETA: 3334s - loss: 11.2893 - acc: 0.025 - ETA: 3361s - loss: 10.2204 - acc: 0.020 - ETA: 3407s - loss: 9.4286 - acc: 0.016 - ETA: 3486s - loss: 8.8228 - acc: 0.01 - ETA: 3464s - loss: 8.3212 - acc: 0.01 - ETA: 3376s - loss: 7.9780 - acc: 0.01 - ETA: 3430s - loss: 7.6717 - acc: 0.01 - ETA: 3358s - loss: 7.4416 - acc: 0.00 - ETA: 3345s - loss: 7.2374 - acc: 0.00 - ETA: 3340s - loss: 7.0555 - acc: 0.00 - ETA: 3303s - loss: 6.8961 - acc: 0.00 - ETA: 3263s - loss: 6.7625 - acc: 0.00 - ETA: 3206s - loss: 6.6492 - acc: 0.00 - ETA: 3127s - loss: 6.5414 - acc: 0.00 - ETA: 3076s - loss: 6.4561 - acc: 0.00 - ETA: 3026s - loss: 6.3737 - acc: 0.00 - ETA: 2971s - loss: 6.3012 - acc: 0.00 - ETA: 2931s - loss: 6.2394 - acc: 0.00 - ETA: 2906s - loss: 6.1772 - acc: 0.00 - ETA: 2976s - loss: 6.1218 - acc: 0.00 - ETA: 2965s - loss: 6.0719 - acc: 0.00 - ETA: 2936s - loss: 6.0257 - acc: 0.00 - ETA: 2912s - loss: 5.9820 - acc: 0.00 - ETA: 2879s - loss: 5.9424 - acc: 0.00 - ETA: 2846s - loss: 5.9044 - acc: 0.01 - ETA: 2828s - loss: 5.8709 - acc: 0.01 - ETA: 2816s - loss: 5.8384 - acc: 0.01 - ETA: 2806s - loss: 5.8078 - acc: 0.00 - ETA: 2788s - loss: 5.7789 - acc: 0.00 - ETA: 2760s - loss: 5.7523 - acc: 0.00 - ETA: 2735s - loss: 5.7270 - acc: 0.00 - ETA: 2714s - loss: 5.7030 - acc: 0.01 - ETA: 2701s - loss: 5.6807 - acc: 0.00 - ETA: 2675s - loss: 5.6592 - acc: 0.00 - ETA: 2661s - loss: 5.6387 - acc: 0.00 - ETA: 2646s - loss: 5.6189 - acc: 0.00 - ETA: 2623s - loss: 5.6004 - acc: 0.01 - ETA: 2610s - loss: 5.5836 - acc: 0.00 - ETA: 2590s - loss: 5.5663 - acc: 0.00 - ETA: 2567s - loss: 5.5497 - acc: 0.00 - ETA: 2551s - loss: 5.5349 - acc: 0.00 - ETA: 2555s - loss: 5.5220 - acc: 0.00 - ETA: 2549s - loss: 5.5078 - acc: 0.00 - ETA: 2534s - loss: 5.4948 - acc: 0.00 - ETA: 2517s - loss: 5.4835 - acc: 0.00 - ETA: 2509s - loss: 5.4714 - acc: 0.00 - ETA: 2498s - loss: 5.4595 - acc: 0.01 - ETA: 2483s - loss: 5.4484 - acc: 0.00 - ETA: 2480s - loss: 5.4376 - acc: 0.01 - ETA: 2492s - loss: 5.4270 - acc: 0.01 - ETA: 2486s - loss: 5.4166 - acc: 0.01 - ETA: 2484s - loss: 5.4070 - acc: 0.01 - ETA: 2467s - loss: 5.3983 - acc: 0.00 - ETA: 2446s - loss: 5.3894 - acc: 0.01 - ETA: 2428s - loss: 5.3809 - acc: 0.01 - ETA: 2406s - loss: 5.3726 - acc: 0.01 - ETA: 2390s - loss: 5.3647 - acc: 0.01 - ETA: 2368s - loss: 5.3569 - acc: 0.00 - ETA: 2348s - loss: 5.3491 - acc: 0.00 - ETA: 2329s - loss: 5.3417 - acc: 0.00 - ETA: 2304s - loss: 5.3342 - acc: 0.01 - ETA: 2287s - loss: 5.3274 - acc: 0.01 - ETA: 2274s - loss: 5.3206 - acc: 0.00 - ETA: 2265s - loss: 5.3134 - acc: 0.01 - ETA: 2250s - loss: 5.3085 - acc: 0.01 - ETA: 2235s - loss: 5.3018 - acc: 0.01 - ETA: 2218s - loss: 5.2963 - acc: 0.01 - ETA: 2206s - loss: 5.2906 - acc: 0.00 - ETA: 2187s - loss: 5.2846 - acc: 0.00 - ETA: 2170s - loss: 5.2790 - acc: 0.00 - ETA: 2154s - loss: 5.2739 - acc: 0.00 - ETA: 2138s - loss: 5.2694 - acc: 0.01 - ETA: 2124s - loss: 5.2640 - acc: 0.01 - ETA: 2111s - loss: 5.2591 - acc: 0.01 - ETA: 2097s - loss: 5.2550 - acc: 0.01 - ETA: 2083s - loss: 5.2502 - acc: 0.01 - ETA: 2066s - loss: 5.2455 - acc: 0.01 - ETA: 2049s - loss: 5.2413 - acc: 0.01 - ETA: 2033s - loss: 5.2369 - acc: 0.01 - ETA: 2017s - loss: 5.2322 - acc: 0.01 - ETA: 2001s - loss: 5.2280 - acc: 0.01 - ETA: 1983s - loss: 5.2236 - acc: 0.01 - ETA: 1971s - loss: 5.2195 - acc: 0.00 - ETA: 1954s - loss: 5.2139 - acc: 0.00 - ETA: 1940s - loss: 5.2123 - acc: 0.00 - ETA: 1926s - loss: 5.2094 - acc: 0.00 - ETA: 1917s - loss: 5.2057 - acc: 0.00 - ETA: 1903s - loss: 5.2026 - acc: 0.00 - ETA: 1889s - loss: 5.1996 - acc: 0.00 - ETA: 1876s - loss: 5.1961 - acc: 0.00 - ETA: 1861s - loss: 5.1929 - acc: 0.00 - ETA: 1849s - loss: 5.1899 - acc: 0.00 - ETA: 1841s - loss: 5.1869 - acc: 0.00 - ETA: 1829s - loss: 5.1840 - acc: 0.00 - ETA: 1819s - loss: 5.1809 - acc: 0.00 - ETA: 1806s - loss: 5.1779 - acc: 0.00 - ETA: 1795s - loss: 5.1751 - acc: 0.00 - ETA: 1783s - loss: 5.1723 - acc: 0.00 - ETA: 1772s - loss: 5.1695 - acc: 0.00 - ETA: 1758s - loss: 5.1672 - acc: 0.00 - ETA: 1748s - loss: 5.1644 - acc: 0.00 - ETA: 1736s - loss: 5.1616 - acc: 0.00 - ETA: 1725s - loss: 5.1589 - acc: 0.00 - ETA: 1715s - loss: 5.1564 - acc: 0.00 - ETA: 1708s - loss: 5.1537 - acc: 0.00 - ETA: 1697s - loss: 5.1518 - acc: 0.00 - ETA: 1685s - loss: 5.1488 - acc: 0.01 - ETA: 1678s - loss: 5.1474 - acc: 0.00 - ETA: 1670s - loss: 5.1449 - acc: 0.00 - ETA: 1666s - loss: 5.1430 - acc: 0.01 - ETA: 1669s - loss: 5.1406 - acc: 0.01 - ETA: 1662s - loss: 5.1383 - acc: 0.01 - ETA: 1654s - loss: 5.1360 - acc: 0.00 - ETA: 1645s - loss: 5.1337 - acc: 0.00 - ETA: 1636s - loss: 5.1318 - acc: 0.00 - ETA: 1627s - loss: 5.1301 - acc: 0.00 - ETA: 1617s - loss: 5.1283 - acc: 0.00 - ETA: 1607s - loss: 5.1260 - acc: 0.00 - ETA: 1599s - loss: 5.1238 - acc: 0.00 - ETA: 1590s - loss: 5.1216 - acc: 0.00 - ETA: 1583s - loss: 5.1201 - acc: 0.00 - ETA: 1575s - loss: 5.1178 - acc: 0.00 - ETA: 1565s - loss: 5.1158 - acc: 0.00 - ETA: 1556s - loss: 5.1137 - acc: 0.00 - ETA: 1545s - loss: 5.1119 - acc: 0.01 - ETA: 1535s - loss: 5.1099 - acc: 0.01 - ETA: 1524s - loss: 5.1078 - acc: 0.01 - ETA: 1516s - loss: 5.1062 - acc: 0.01 - ETA: 1507s - loss: 5.1044 - acc: 0.01 - ETA: 1496s - loss: 5.1026 - acc: 0.01 - ETA: 1487s - loss: 5.1010 - acc: 0.01 - ETA: 1477s - loss: 5.0990 - acc: 0.01 - ETA: 1468s - loss: 5.0968 - acc: 0.01 - ETA: 1458s - loss: 5.0956 - acc: 0.01 - ETA: 1450s - loss: 5.0942 - acc: 0.01 - ETA: 1443s - loss: 5.0927 - acc: 0.01 - ETA: 1436s - loss: 5.0912 - acc: 0.01 - ETA: 1427s - loss: 5.0896 - acc: 0.00 - ETA: 1419s - loss: 5.0879 - acc: 0.00 - ETA: 1411s - loss: 5.0851 - acc: 0.01 - ETA: 1402s - loss: 5.0854 - acc: 0.01 - ETA: 1396s - loss: 5.0837 - acc: 0.01 - ETA: 1388s - loss: 5.0821 - acc: 0.00 - ETA: 1379s - loss: 5.0800 - acc: 0.00 - ETA: 1370s - loss: 5.0783 - acc: 0.00 - ETA: 1364s - loss: 5.0775 - acc: 0.00 - ETA: 1355s - loss: 5.0763 - acc: 0.00 - ETA: 1347s - loss: 5.0747 - acc: 0.00 - ETA: 1340s - loss: 5.0737 - acc: 0.00 - ETA: 1331s - loss: 5.0721 - acc: 0.00 - ETA: 1323s - loss: 5.0706 - acc: 0.01 - ETA: 1315s - loss: 5.0692 - acc: 0.01 - ETA: 1307s - loss: 5.0682 - acc: 0.00 - ETA: 1298s - loss: 5.0673 - acc: 0.01 - ETA: 1291s - loss: 5.0659 - acc: 0.01 - ETA: 1283s - loss: 5.0646 - acc: 0.01 - ETA: 1275s - loss: 5.0635 - acc: 0.01 - ETA: 1265s - loss: 5.0624 - acc: 0.00 - ETA: 1257s - loss: 5.0611 - acc: 0.00 - ETA: 1248s - loss: 5.0600 - acc: 0.00 - ETA: 1240s - loss: 5.0590 - acc: 0.00 - ETA: 1232s - loss: 5.0582 - acc: 0.00 - ETA: 1223s - loss: 5.0568 - acc: 0.00 - ETA: 1214s - loss: 5.0555 - acc: 0.00 - ETA: 1206s - loss: 5.0544 - acc: 0.00 - ETA: 1198s - loss: 5.0535 - acc: 0.00 - ETA: 1190s - loss: 5.0520 - acc: 0.00 - ETA: 1182s - loss: 5.0508 - acc: 0.00 - ETA: 1174s - loss: 5.0489 - acc: 0.00 - ETA: 1166s - loss: 5.0479 - acc: 0.00 - ETA: 1157s - loss: 5.0466 - acc: 0.01 - ETA: 1149s - loss: 5.0459 - acc: 0.01 - ETA: 1141s - loss: 5.0444 - acc: 0.01 - ETA: 1134s - loss: 5.0425 - acc: 0.01 - ETA: 1128s - loss: 5.0415 - acc: 0.01 - ETA: 1120s - loss: 5.0406 - acc: 0.01 - ETA: 1112s - loss: 5.0394 - acc: 0.01 - ETA: 1104s - loss: 5.0378 - acc: 0.01 - ETA: 1096s - loss: 5.0364 - acc: 0.01 - ETA: 1087s - loss: 5.0351 - acc: 0.01 - ETA: 1080s - loss: 5.0337 - acc: 0.01 - ETA: 1075s - loss: 5.0327 - acc: 0.01 - ETA: 1068s - loss: 5.0312 - acc: 0.01 - ETA: 1061s - loss: 5.0297 - acc: 0.01 - ETA: 1054s - loss: 5.0288 - acc: 0.01 - ETA: 1048s - loss: 5.0286 - acc: 0.01 - ETA: 1040s - loss: 5.0286 - acc: 0.01 - ETA: 1032s - loss: 5.0269 - acc: 0.01 - ETA: 1026s - loss: 5.0269 - acc: 0.01 - ETA: 1018s - loss: 5.0257 - acc: 0.01 - ETA: 1012s - loss: 5.0243 - acc: 0.01 - ETA: 1004s - loss: 5.0229 - acc: 0.01 - ETA: 996s - loss: 5.0222 - acc: 0.0110 - ETA: 988s - loss: 5.0209 - acc: 0.011 - ETA: 980s - loss: 5.0201 - acc: 0.011 - ETA: 972s - loss: 5.0195 - acc: 0.011 - ETA: 965s - loss: 5.0189 - acc: 0.011 - ETA: 956s - loss: 5.0178 - acc: 0.011 - ETA: 948s - loss: 5.0162 - acc: 0.01166660/6680 [============================>.] - ETA: 940s - loss: 5.0153 - acc: 0.011 - ETA: 932s - loss: 5.0135 - acc: 0.012 - ETA: 925s - loss: 5.0138 - acc: 0.012 - ETA: 917s - loss: 5.0133 - acc: 0.011 - ETA: 909s - loss: 5.0121 - acc: 0.012 - ETA: 901s - loss: 5.0110 - acc: 0.012 - ETA: 893s - loss: 5.0109 - acc: 0.012 - ETA: 885s - loss: 5.0105 - acc: 0.012 - ETA: 878s - loss: 5.0093 - acc: 0.012 - ETA: 870s - loss: 5.0087 - acc: 0.012 - ETA: 863s - loss: 5.0072 - acc: 0.012 - ETA: 855s - loss: 5.0070 - acc: 0.012 - ETA: 847s - loss: 5.0064 - acc: 0.012 - ETA: 838s - loss: 5.0053 - acc: 0.012 - ETA: 830s - loss: 5.0047 - acc: 0.012 - ETA: 821s - loss: 5.0040 - acc: 0.012 - ETA: 812s - loss: 5.0025 - acc: 0.012 - ETA: 804s - loss: 5.0013 - acc: 0.012 - ETA: 795s - loss: 4.9995 - acc: 0.012 - ETA: 786s - loss: 4.9986 - acc: 0.012 - ETA: 778s - loss: 4.9980 - acc: 0.012 - ETA: 770s - loss: 4.9973 - acc: 0.012 - ETA: 761s - loss: 4.9960 - acc: 0.012 - ETA: 753s - loss: 4.9945 - acc: 0.013 - ETA: 745s - loss: 4.9924 - acc: 0.013 - ETA: 736s - loss: 4.9915 - acc: 0.012 - ETA: 728s - loss: 4.9894 - acc: 0.013 - ETA: 720s - loss: 4.9884 - acc: 0.013 - ETA: 712s - loss: 4.9870 - acc: 0.013 - ETA: 703s - loss: 4.9861 - acc: 0.013 - ETA: 695s - loss: 4.9863 - acc: 0.013 - ETA: 687s - loss: 4.9859 - acc: 0.013 - ETA: 679s - loss: 4.9850 - acc: 0.013 - ETA: 671s - loss: 4.9838 - acc: 0.012 - ETA: 664s - loss: 4.9831 - acc: 0.012 - ETA: 656s - loss: 4.9830 - acc: 0.012 - ETA: 648s - loss: 4.9822 - acc: 0.013 - ETA: 641s - loss: 4.9814 - acc: 0.013 - ETA: 633s - loss: 4.9801 - acc: 0.013 - ETA: 625s - loss: 4.9792 - acc: 0.013 - ETA: 617s - loss: 4.9780 - acc: 0.013 - ETA: 610s - loss: 4.9774 - acc: 0.013 - ETA: 602s - loss: 4.9767 - acc: 0.013 - ETA: 593s - loss: 4.9750 - acc: 0.013 - ETA: 586s - loss: 4.9740 - acc: 0.013 - ETA: 578s - loss: 4.9734 - acc: 0.013 - ETA: 571s - loss: 4.9709 - acc: 0.014 - ETA: 563s - loss: 4.9709 - acc: 0.014 - ETA: 555s - loss: 4.9687 - acc: 0.014 - ETA: 548s - loss: 4.9685 - acc: 0.014 - ETA: 540s - loss: 4.9679 - acc: 0.014 - ETA: 533s - loss: 4.9665 - acc: 0.014 - ETA: 525s - loss: 4.9658 - acc: 0.014 - ETA: 518s - loss: 4.9649 - acc: 0.014 - ETA: 511s - loss: 4.9637 - acc: 0.014 - ETA: 503s - loss: 4.9629 - acc: 0.014 - ETA: 496s - loss: 4.9625 - acc: 0.014 - ETA: 489s - loss: 4.9623 - acc: 0.014 - ETA: 482s - loss: 4.9615 - acc: 0.014 - ETA: 475s - loss: 4.9604 - acc: 0.014 - ETA: 467s - loss: 4.9594 - acc: 0.014 - ETA: 460s - loss: 4.9587 - acc: 0.014 - ETA: 453s - loss: 4.9562 - acc: 0.015 - ETA: 446s - loss: 4.9559 - acc: 0.015 - ETA: 438s - loss: 4.9540 - acc: 0.015 - ETA: 431s - loss: 4.9527 - acc: 0.015 - ETA: 424s - loss: 4.9522 - acc: 0.015 - ETA: 417s - loss: 4.9506 - acc: 0.015 - ETA: 410s - loss: 4.9498 - acc: 0.015 - ETA: 403s - loss: 4.9493 - acc: 0.015 - ETA: 396s - loss: 4.9488 - acc: 0.015 - ETA: 388s - loss: 4.9479 - acc: 0.015 - ETA: 381s - loss: 4.9463 - acc: 0.015 - ETA: 374s - loss: 4.9461 - acc: 0.015 - ETA: 367s - loss: 4.9442 - acc: 0.016 - ETA: 361s - loss: 4.9425 - acc: 0.016 - ETA: 354s - loss: 4.9422 - acc: 0.016 - ETA: 347s - loss: 4.9423 - acc: 0.016 - ETA: 340s - loss: 4.9422 - acc: 0.016 - ETA: 333s - loss: 4.9420 - acc: 0.016 - ETA: 326s - loss: 4.9416 - acc: 0.015 - ETA: 319s - loss: 4.9408 - acc: 0.016 - ETA: 312s - loss: 4.9401 - acc: 0.016 - ETA: 306s - loss: 4.9397 - acc: 0.015 - ETA: 299s - loss: 4.9397 - acc: 0.015 - ETA: 292s - loss: 4.9382 - acc: 0.016 - ETA: 285s - loss: 4.9366 - acc: 0.016 - ETA: 279s - loss: 4.9353 - acc: 0.016 - ETA: 272s - loss: 4.9337 - acc: 0.016 - ETA: 265s - loss: 4.9327 - acc: 0.016 - ETA: 258s - loss: 4.9319 - acc: 0.016 - ETA: 252s - loss: 4.9301 - acc: 0.017 - ETA: 245s - loss: 4.9301 - acc: 0.016 - ETA: 239s - loss: 4.9295 - acc: 0.016 - ETA: 232s - loss: 4.9283 - acc: 0.017 - ETA: 226s - loss: 4.9274 - acc: 0.017 - ETA: 219s - loss: 4.9260 - acc: 0.017 - ETA: 213s - loss: 4.9261 - acc: 0.017 - ETA: 206s - loss: 4.9262 - acc: 0.017 - ETA: 200s - loss: 4.9262 - acc: 0.017 - ETA: 193s - loss: 4.9253 - acc: 0.017 - ETA: 187s - loss: 4.9239 - acc: 0.017 - ETA: 180s - loss: 4.9227 - acc: 0.017 - ETA: 174s - loss: 4.9227 - acc: 0.017 - ETA: 168s - loss: 4.9226 - acc: 0.017 - ETA: 161s - loss: 4.9214 - acc: 0.017 - ETA: 155s - loss: 4.9211 - acc: 0.017 - ETA: 148s - loss: 4.9211 - acc: 0.017 - ETA: 142s - loss: 4.9202 - acc: 0.018 - ETA: 136s - loss: 4.9196 - acc: 0.017 - ETA: 129s - loss: 4.9189 - acc: 0.017 - ETA: 123s - loss: 4.9176 - acc: 0.018 - ETA: 117s - loss: 4.9172 - acc: 0.017 - ETA: 111s - loss: 4.9161 - acc: 0.018 - ETA: 104s - loss: 4.9141 - acc: 0.018 - ETA: 98s - loss: 4.9134 - acc: 0.018 - ETA: 92s - loss: 4.9116 - acc: 0.01 - ETA: 85s - loss: 4.9107 - acc: 0.01 - ETA: 79s - loss: 4.9099 - acc: 0.01 - ETA: 73s - loss: 4.9090 - acc: 0.01 - ETA: 67s - loss: 4.9082 - acc: 0.01 - ETA: 61s - loss: 4.9074 - acc: 0.01 - ETA: 54s - loss: 4.9072 - acc: 0.01 - ETA: 48s - loss: 4.9063 - acc: 0.01 - ETA: 42s - loss: 4.9065 - acc: 0.01 - ETA: 36s - loss: 4.9057 - acc: 0.01 - ETA: 30s - loss: 4.9043 - acc: 0.01 - ETA: 24s - loss: 4.9037 - acc: 0.01 - ETA: 18s - loss: 4.9031 - acc: 0.01 - ETA: 12s - loss: 4.9021 - acc: 0.01 - ETA: 6s - loss: 4.9014 - acc: 0.0192 Epoch 00000: val_loss improved from inf to 4.55192, saving model to saved_models/weights.best.from_scratch.hdf5 6680/6680 [==============================] - 2046s - loss: 4.9003 - acc: 0.0192 - val_loss: 4.5519 - val_acc: 0.0419 Epoch 2/5 4080/6680 [=================>............] - ETA: 1794s - loss: 4.5485 - acc: 0.10 - ETA: 1667s - loss: 4.6678 - acc: 0.05 - ETA: 1579s - loss: 4.6351 - acc: 0.06 - ETA: 1522s - loss: 4.5510 - acc: 0.06 - ETA: 1473s - loss: 4.5448 - acc: 0.05 - ETA: 1455s - loss: 4.5059 - acc: 0.05 - ETA: 1435s - loss: 4.5227 - acc: 0.05 - ETA: 1422s - loss: 4.5007 - acc: 0.05 - ETA: 1398s - loss: 4.4544 - acc: 0.05 - ETA: 1373s - loss: 4.4631 - acc: 0.04 - ETA: 1383s - loss: 4.4913 - acc: 0.04 - ETA: 1364s - loss: 4.4812 - acc: 0.05 - ETA: 1352s - loss: 4.4865 - acc: 0.04 - ETA: 1356s - loss: 4.4658 - acc: 0.05 - ETA: 1340s - loss: 4.4779 - acc: 0.04 - ETA: 1327s - loss: 4.4843 - acc: 0.04 - ETA: 1335s - loss: 4.4681 - acc: 0.05 - ETA: 1337s - loss: 4.4570 - acc: 0.05 - ETA: 1338s - loss: 4.4484 - acc: 0.05 - ETA: 1317s - loss: 4.4572 - acc: 0.05 - ETA: 1317s - loss: 4.4732 - acc: 0.04 - ETA: 1330s - loss: 4.4711 - acc: 0.05 - ETA: 1381s - loss: 4.4529 - acc: 0.05 - ETA: 1431s - loss: 4.4427 - acc: 0.05 - ETA: 1479s - loss: 4.4383 - acc: 0.05 - ETA: 1526s - loss: 4.4388 - acc: 0.05 - ETA: 1558s - loss: 4.4275 - acc: 0.05 - ETA: 1572s - loss: 4.4346 - acc: 0.05 - ETA: 1596s - loss: 4.4365 - acc: 0.05 - ETA: 1599s - loss: 4.4490 - acc: 0.05 - ETA: 1601s - loss: 4.4585 - acc: 0.05 - ETA: 1613s - loss: 4.4562 - acc: 0.05 - ETA: 1617s - loss: 4.4653 - acc: 0.05 - ETA: 1619s - loss: 4.4721 - acc: 0.05 - ETA: 1648s - loss: 4.4721 - acc: 0.05 - ETA: 1655s - loss: 4.4689 - acc: 0.05 - ETA: 1671s - loss: 4.4703 - acc: 0.05 - ETA: 1681s - loss: 4.4719 - acc: 0.05 - ETA: 1690s - loss: 4.4749 - acc: 0.05 - ETA: 1691s - loss: 4.4712 - acc: 0.05 - ETA: 1689s - loss: 4.4704 - acc: 0.05 - ETA: 1688s - loss: 4.4730 - acc: 0.05 - ETA: 1681s - loss: 4.4712 - acc: 0.05 - ETA: 1685s - loss: 4.4727 - acc: 0.04 - ETA: 1694s - loss: 4.4700 - acc: 0.04 - ETA: 1700s - loss: 4.4666 - acc: 0.04 - ETA: 1698s - loss: 4.4676 - acc: 0.04 - ETA: 1698s - loss: 4.4589 - acc: 0.05 - ETA: 1694s - loss: 4.4514 - acc: 0.05 - ETA: 1695s - loss: 4.4659 - acc: 0.04 - ETA: 1691s - loss: 4.4738 - acc: 0.04 - ETA: 1692s - loss: 4.4810 - acc: 0.04 - ETA: 1693s - loss: 4.4779 - acc: 0.04 - ETA: 1685s - loss: 4.4742 - acc: 0.04 - ETA: 1681s - loss: 4.4747 - acc: 0.04 - ETA: 1694s - loss: 4.4724 - acc: 0.04 - ETA: 1698s - loss: 4.4771 - acc: 0.04 - ETA: 1703s - loss: 4.4800 - acc: 0.04 - ETA: 1699s - loss: 4.4793 - acc: 0.04 - ETA: 1691s - loss: 4.4784 - acc: 0.04 - ETA: 1688s - loss: 4.4791 - acc: 0.04 - ETA: 1688s - loss: 4.4789 - acc: 0.04 - ETA: 1680s - loss: 4.4742 - acc: 0.04 - ETA: 1684s - loss: 4.4720 - acc: 0.04 - ETA: 1683s - loss: 4.4685 - acc: 0.04 - ETA: 1674s - loss: 4.4667 - acc: 0.04 - ETA: 1668s - loss: 4.4714 - acc: 0.04 - ETA: 1661s - loss: 4.4667 - acc: 0.05 - ETA: 1655s - loss: 4.4656 - acc: 0.05 - ETA: 1647s - loss: 4.4651 - acc: 0.05 - ETA: 1638s - loss: 4.4615 - acc: 0.05 - ETA: 1631s - loss: 4.4618 - acc: 0.05 - ETA: 1625s - loss: 4.4593 - acc: 0.05 - ETA: 1621s - loss: 4.4589 - acc: 0.05 - ETA: 1616s - loss: 4.4619 - acc: 0.05 - ETA: 1608s - loss: 4.4612 - acc: 0.05 - ETA: 1602s - loss: 4.4588 - acc: 0.05 - ETA: 1592s - loss: 4.4559 - acc: 0.05 - ETA: 1587s - loss: 4.4557 - acc: 0.05 - ETA: 1581s - loss: 4.4544 - acc: 0.05 - ETA: 1571s - loss: 4.4577 - acc: 0.05 - ETA: 1563s - loss: 4.4566 - acc: 0.05 - ETA: 1555s - loss: 4.4546 - acc: 0.05 - ETA: 1548s - loss: 4.4564 - acc: 0.05 - ETA: 1541s - loss: 4.4556 - acc: 0.05 - ETA: 1534s - loss: 4.4540 - acc: 0.05 - ETA: 1531s - loss: 4.4519 - acc: 0.05 - ETA: 1528s - loss: 4.4537 - acc: 0.05 - ETA: 1523s - loss: 4.4558 - acc: 0.05 - ETA: 1519s - loss: 4.4572 - acc: 0.05 - ETA: 1515s - loss: 4.4552 - acc: 0.05 - ETA: 1509s - loss: 4.4613 - acc: 0.05 - ETA: 1500s - loss: 4.4598 - acc: 0.05 - ETA: 1491s - loss: 4.4617 - acc: 0.05 - ETA: 1484s - loss: 4.4631 - acc: 0.05 - ETA: 1476s - loss: 4.4614 - acc: 0.05 - ETA: 1469s - loss: 4.4603 - acc: 0.05 - ETA: 1463s - loss: 4.4636 - acc: 0.04 - ETA: 1458s - loss: 4.4631 - acc: 0.04 - ETA: 1457s - loss: 4.4634 - acc: 0.04 - ETA: 1449s - loss: 4.4626 - acc: 0.04 - ETA: 1443s - loss: 4.4610 - acc: 0.04 - ETA: 1435s - loss: 4.4606 - acc: 0.04 - ETA: 1428s - loss: 4.4613 - acc: 0.04 - ETA: 1421s - loss: 4.4605 - acc: 0.04 - ETA: 1414s - loss: 4.4634 - acc: 0.04 - ETA: 1407s - loss: 4.4638 - acc: 0.04 - ETA: 1401s - loss: 4.4618 - acc: 0.04 - ETA: 1396s - loss: 4.4607 - acc: 0.04 - ETA: 1391s - loss: 4.4581 - acc: 0.04 - ETA: 1387s - loss: 4.4570 - acc: 0.04 - ETA: 1383s - loss: 4.4547 - acc: 0.04 - ETA: 1376s - loss: 4.4591 - acc: 0.04 - ETA: 1370s - loss: 4.4572 - acc: 0.04 - ETA: 1363s - loss: 4.4565 - acc: 0.04 - ETA: 1355s - loss: 4.4541 - acc: 0.04 - ETA: 1347s - loss: 4.4548 - acc: 0.04 - ETA: 1341s - loss: 4.4559 - acc: 0.04 - ETA: 1334s - loss: 4.4568 - acc: 0.04 - ETA: 1328s - loss: 4.4559 - acc: 0.04 - ETA: 1324s - loss: 4.4556 - acc: 0.04 - ETA: 1318s - loss: 4.4560 - acc: 0.04 - ETA: 1311s - loss: 4.4556 - acc: 0.04 - ETA: 1305s - loss: 4.4544 - acc: 0.04 - ETA: 1300s - loss: 4.4539 - acc: 0.04 - ETA: 1293s - loss: 4.4561 - acc: 0.04 - ETA: 1286s - loss: 4.4568 - acc: 0.04 - ETA: 1280s - loss: 4.4547 - acc: 0.04 - ETA: 1273s - loss: 4.4535 - acc: 0.04 - ETA: 1267s - loss: 4.4532 - acc: 0.04 - ETA: 1261s - loss: 4.4535 - acc: 0.04 - ETA: 1255s - loss: 4.4513 - acc: 0.04 - ETA: 1249s - loss: 4.4524 - acc: 0.04 - ETA: 1243s - loss: 4.4506 - acc: 0.04 - ETA: 1237s - loss: 4.4448 - acc: 0.04 - ETA: 1231s - loss: 4.4457 - acc: 0.04 - ETA: 1226s - loss: 4.4441 - acc: 0.04 - ETA: 1219s - loss: 4.4452 - acc: 0.04 - ETA: 1211s - loss: 4.4454 - acc: 0.04 - ETA: 1204s - loss: 4.4437 - acc: 0.04 - ETA: 1196s - loss: 4.4441 - acc: 0.04 - ETA: 1187s - loss: 4.4459 - acc: 0.04 - ETA: 1178s - loss: 4.4444 - acc: 0.04 - ETA: 1169s - loss: 4.4433 - acc: 0.04 - ETA: 1160s - loss: 4.4424 - acc: 0.04 - ETA: 1151s - loss: 4.4395 - acc: 0.04 - ETA: 1141s - loss: 4.4414 - acc: 0.05 - ETA: 1132s - loss: 4.4419 - acc: 0.04 - ETA: 1124s - loss: 4.4417 - acc: 0.04 - ETA: 1115s - loss: 4.4386 - acc: 0.04 - ETA: 1107s - loss: 4.4369 - acc: 0.05 - ETA: 1098s - loss: 4.4357 - acc: 0.04 - ETA: 1091s - loss: 4.4353 - acc: 0.04 - ETA: 1084s - loss: 4.4348 - acc: 0.04 - ETA: 1076s - loss: 4.4359 - acc: 0.05 - ETA: 1068s - loss: 4.4369 - acc: 0.05 - ETA: 1060s - loss: 4.4359 - acc: 0.05 - ETA: 1051s - loss: 4.4361 - acc: 0.04 - ETA: 1043s - loss: 4.4377 - acc: 0.04 - ETA: 1035s - loss: 4.4395 - acc: 0.04 - ETA: 1028s - loss: 4.4386 - acc: 0.05 - ETA: 1019s - loss: 4.4370 - acc: 0.04 - ETA: 1011s - loss: 4.4339 - acc: 0.05 - ETA: 1004s - loss: 4.4344 - acc: 0.05 - ETA: 997s - loss: 4.4386 - acc: 0.0503 - ETA: 989s - loss: 4.4385 - acc: 0.050 - ETA: 981s - loss: 4.4377 - acc: 0.050 - ETA: 973s - loss: 4.4380 - acc: 0.050 - ETA: 966s - loss: 4.4363 - acc: 0.050 - ETA: 958s - loss: 4.4349 - acc: 0.050 - ETA: 950s - loss: 4.4332 - acc: 0.051 - ETA: 943s - loss: 4.4331 - acc: 0.050 - ETA: 936s - loss: 4.4321 - acc: 0.051 - ETA: 929s - loss: 4.4310 - acc: 0.051 - ETA: 922s - loss: 4.4311 - acc: 0.050 - ETA: 914s - loss: 4.4296 - acc: 0.050 - ETA: 907s - loss: 4.4271 - acc: 0.050 - ETA: 900s - loss: 4.4261 - acc: 0.050 - ETA: 894s - loss: 4.4261 - acc: 0.050 - ETA: 887s - loss: 4.4249 - acc: 0.050 - ETA: 880s - loss: 4.4243 - acc: 0.050 - ETA: 873s - loss: 4.4242 - acc: 0.050 - ETA: 867s - loss: 4.4242 - acc: 0.050 - ETA: 860s - loss: 4.4221 - acc: 0.050 - ETA: 853s - loss: 4.4210 - acc: 0.050 - ETA: 846s - loss: 4.4202 - acc: 0.050 - ETA: 839s - loss: 4.4192 - acc: 0.050 - ETA: 832s - loss: 4.4170 - acc: 0.050 - ETA: 825s - loss: 4.4160 - acc: 0.050 - ETA: 818s - loss: 4.4157 - acc: 0.051 - ETA: 811s - loss: 4.4159 - acc: 0.050 - ETA: 804s - loss: 4.4149 - acc: 0.050 - ETA: 798s - loss: 4.4186 - acc: 0.050 - ETA: 792s - loss: 4.4176 - acc: 0.050 - ETA: 786s - loss: 4.4187 - acc: 0.050 - ETA: 779s - loss: 4.4194 - acc: 0.050 - ETA: 773s - loss: 4.4195 - acc: 0.050 - ETA: 767s - loss: 4.4190 - acc: 0.050 - ETA: 760s - loss: 4.4182 - acc: 0.050 - ETA: 754s - loss: 4.4171 - acc: 0.050 - ETA: 747s - loss: 4.4164 - acc: 0.050 - ETA: 741s - loss: 4.4175 - acc: 0.050 - ETA: 734s - loss: 4.4153 - acc: 0.050 - ETA: 727s - loss: 4.4148 - acc: 0.05006660/6680 [============================>.] - ETA: 720s - loss: 4.4154 - acc: 0.049 - ETA: 714s - loss: 4.4167 - acc: 0.049 - ETA: 708s - loss: 4.4189 - acc: 0.049 - ETA: 701s - loss: 4.4194 - acc: 0.049 - ETA: 695s - loss: 4.4199 - acc: 0.049 - ETA: 689s - loss: 4.4191 - acc: 0.049 - ETA: 682s - loss: 4.4183 - acc: 0.049 - ETA: 677s - loss: 4.4182 - acc: 0.049 - ETA: 670s - loss: 4.4167 - acc: 0.049 - ETA: 664s - loss: 4.4163 - acc: 0.049 - ETA: 658s - loss: 4.4161 - acc: 0.049 - ETA: 652s - loss: 4.4157 - acc: 0.049 - ETA: 645s - loss: 4.4152 - acc: 0.049 - ETA: 639s - loss: 4.4127 - acc: 0.050 - ETA: 633s - loss: 4.4103 - acc: 0.050 - ETA: 627s - loss: 4.4093 - acc: 0.050 - ETA: 621s - loss: 4.4091 - acc: 0.050 - ETA: 615s - loss: 4.4077 - acc: 0.050 - ETA: 609s - loss: 4.4064 - acc: 0.050 - ETA: 603s - loss: 4.4060 - acc: 0.050 - ETA: 597s - loss: 4.4062 - acc: 0.050 - ETA: 591s - loss: 4.4054 - acc: 0.050 - ETA: 585s - loss: 4.4072 - acc: 0.050 - ETA: 579s - loss: 4.4065 - acc: 0.050 - ETA: 573s - loss: 4.4062 - acc: 0.050 - ETA: 566s - loss: 4.4056 - acc: 0.051 - ETA: 560s - loss: 4.4055 - acc: 0.051 - ETA: 554s - loss: 4.4052 - acc: 0.050 - ETA: 548s - loss: 4.4046 - acc: 0.050 - ETA: 543s - loss: 4.4047 - acc: 0.051 - ETA: 537s - loss: 4.4054 - acc: 0.051 - ETA: 531s - loss: 4.4047 - acc: 0.051 - ETA: 525s - loss: 4.4062 - acc: 0.050 - ETA: 519s - loss: 4.4059 - acc: 0.051 - ETA: 513s - loss: 4.4053 - acc: 0.051 - ETA: 508s - loss: 4.4064 - acc: 0.051 - ETA: 502s - loss: 4.4050 - acc: 0.051 - ETA: 496s - loss: 4.4045 - acc: 0.051 - ETA: 490s - loss: 4.4049 - acc: 0.051 - ETA: 484s - loss: 4.4035 - acc: 0.051 - ETA: 478s - loss: 4.4032 - acc: 0.051 - ETA: 472s - loss: 4.4027 - acc: 0.051 - ETA: 467s - loss: 4.4035 - acc: 0.051 - ETA: 461s - loss: 4.4028 - acc: 0.051 - ETA: 455s - loss: 4.4037 - acc: 0.051 - ETA: 449s - loss: 4.4034 - acc: 0.051 - ETA: 444s - loss: 4.4029 - acc: 0.052 - ETA: 438s - loss: 4.4025 - acc: 0.052 - ETA: 432s - loss: 4.4021 - acc: 0.052 - ETA: 427s - loss: 4.4023 - acc: 0.052 - ETA: 421s - loss: 4.4019 - acc: 0.052 - ETA: 415s - loss: 4.4030 - acc: 0.052 - ETA: 410s - loss: 4.4026 - acc: 0.052 - ETA: 404s - loss: 4.4018 - acc: 0.052 - ETA: 399s - loss: 4.4010 - acc: 0.052 - ETA: 393s - loss: 4.4005 - acc: 0.052 - ETA: 387s - loss: 4.3992 - acc: 0.052 - ETA: 381s - loss: 4.3974 - acc: 0.052 - ETA: 376s - loss: 4.3978 - acc: 0.052 - ETA: 370s - loss: 4.3976 - acc: 0.052 - ETA: 365s - loss: 4.3978 - acc: 0.052 - ETA: 359s - loss: 4.3966 - acc: 0.052 - ETA: 354s - loss: 4.3953 - acc: 0.052 - ETA: 348s - loss: 4.3962 - acc: 0.052 - ETA: 343s - loss: 4.3972 - acc: 0.052 - ETA: 337s - loss: 4.3965 - acc: 0.052 - ETA: 332s - loss: 4.3959 - acc: 0.052 - ETA: 326s - loss: 4.3952 - acc: 0.052 - ETA: 321s - loss: 4.3958 - acc: 0.051 - ETA: 316s - loss: 4.3929 - acc: 0.052 - ETA: 311s - loss: 4.3926 - acc: 0.051 - ETA: 306s - loss: 4.3940 - acc: 0.051 - ETA: 302s - loss: 4.3945 - acc: 0.051 - ETA: 297s - loss: 4.3956 - acc: 0.051 - ETA: 292s - loss: 4.3959 - acc: 0.051 - ETA: 287s - loss: 4.3963 - acc: 0.051 - ETA: 282s - loss: 4.3968 - acc: 0.051 - ETA: 276s - loss: 4.3976 - acc: 0.050 - ETA: 271s - loss: 4.3977 - acc: 0.050 - ETA: 266s - loss: 4.3971 - acc: 0.051 - ETA: 261s - loss: 4.3963 - acc: 0.051 - ETA: 256s - loss: 4.3956 - acc: 0.051 - ETA: 250s - loss: 4.3957 - acc: 0.051 - ETA: 245s - loss: 4.3963 - acc: 0.051 - ETA: 240s - loss: 4.3969 - acc: 0.051 - ETA: 235s - loss: 4.3955 - acc: 0.051 - ETA: 229s - loss: 4.3956 - acc: 0.051 - ETA: 224s - loss: 4.3942 - acc: 0.051 - ETA: 219s - loss: 4.3931 - acc: 0.051 - ETA: 214s - loss: 4.3939 - acc: 0.051 - ETA: 209s - loss: 4.3921 - acc: 0.051 - ETA: 203s - loss: 4.3909 - acc: 0.051 - ETA: 198s - loss: 4.3912 - acc: 0.051 - ETA: 193s - loss: 4.3903 - acc: 0.052 - ETA: 187s - loss: 4.3891 - acc: 0.052 - ETA: 182s - loss: 4.3875 - acc: 0.052 - ETA: 177s - loss: 4.3875 - acc: 0.052 - ETA: 171s - loss: 4.3851 - acc: 0.053 - ETA: 166s - loss: 4.3842 - acc: 0.053 - ETA: 161s - loss: 4.3831 - acc: 0.053 - ETA: 155s - loss: 4.3816 - acc: 0.053 - ETA: 150s - loss: 4.3823 - acc: 0.053 - ETA: 145s - loss: 4.3826 - acc: 0.053 - ETA: 139s - loss: 4.3841 - acc: 0.053 - ETA: 134s - loss: 4.3836 - acc: 0.053 - ETA: 129s - loss: 4.3825 - acc: 0.053 - ETA: 123s - loss: 4.3813 - acc: 0.053 - ETA: 118s - loss: 4.3812 - acc: 0.053 - ETA: 113s - loss: 4.3811 - acc: 0.053 - ETA: 107s - loss: 4.3812 - acc: 0.053 - ETA: 102s - loss: 4.3805 - acc: 0.053 - ETA: 97s - loss: 4.3799 - acc: 0.054 - ETA: 91s - loss: 4.3796 - acc: 0.05 - ETA: 86s - loss: 4.3784 - acc: 0.05 - ETA: 81s - loss: 4.3788 - acc: 0.05 - ETA: 75s - loss: 4.3778 - acc: 0.05 - ETA: 70s - loss: 4.3774 - acc: 0.05 - ETA: 64s - loss: 4.3775 - acc: 0.05 - ETA: 59s - loss: 4.3765 - acc: 0.05 - ETA: 54s - loss: 4.3767 - acc: 0.05 - ETA: 48s - loss: 4.3773 - acc: 0.05 - ETA: 43s - loss: 4.3779 - acc: 0.05 - ETA: 37s - loss: 4.3781 - acc: 0.05 - ETA: 32s - loss: 4.3777 - acc: 0.05 - ETA: 27s - loss: 4.3779 - acc: 0.05 - ETA: 21s - loss: 4.3773 - acc: 0.05 - ETA: 16s - loss: 4.3769 - acc: 0.05 - ETA: 10s - loss: 4.3759 - acc: 0.05 - ETA: 5s - loss: 4.3755 - acc: 0.0539 Epoch 00001: val_loss improved from 4.55192 to 4.29654, saving model to saved_models/weights.best.from_scratch.hdf5 6680/6680 [==============================] - 1836s - loss: 4.3753 - acc: 0.0539 - val_loss: 4.2965 - val_acc: 0.0563 Epoch 3/5 4080/6680 [=================>............] - ETA: 2242s - loss: 4.4395 - acc: 0.0000e+ - ETA: 2218s - loss: 4.1894 - acc: 0.0500 - ETA: 2719s - loss: 4.2797 - acc: 0.08 - ETA: 2844s - loss: 4.2165 - acc: 0.11 - ETA: 2717s - loss: 4.1193 - acc: 0.12 - ETA: 2609s - loss: 4.0511 - acc: 0.12 - ETA: 2531s - loss: 4.3223 - acc: 0.11 - ETA: 2478s - loss: 4.3174 - acc: 0.11 - ETA: 2415s - loss: 4.3105 - acc: 0.11 - ETA: 2374s - loss: 4.2787 - acc: 0.12 - ETA: 2338s - loss: 4.2631 - acc: 0.12 - ETA: 2311s - loss: 4.2274 - acc: 0.12 - ETA: 2270s - loss: 4.2260 - acc: 0.11 - ETA: 2263s - loss: 4.1924 - acc: 0.11 - ETA: 2226s - loss: 4.1973 - acc: 0.10 - ETA: 2217s - loss: 4.1864 - acc: 0.11 - ETA: 2207s - loss: 4.1562 - acc: 0.11 - ETA: 2205s - loss: 4.1600 - acc: 0.11 - ETA: 2207s - loss: 4.1381 - acc: 0.12 - ETA: 2213s - loss: 4.1367 - acc: 0.12 - ETA: 2221s - loss: 4.1291 - acc: 0.12 - ETA: 2239s - loss: 4.1326 - acc: 0.12 - ETA: 2237s - loss: 4.1340 - acc: 0.12 - ETA: 2227s - loss: 4.1300 - acc: 0.12 - ETA: 2222s - loss: 4.1088 - acc: 0.12 - ETA: 2215s - loss: 4.1016 - acc: 0.12 - ETA: 2203s - loss: 4.1137 - acc: 0.12 - ETA: 2190s - loss: 4.0926 - acc: 0.12 - ETA: 2196s - loss: 4.0879 - acc: 0.12 - ETA: 2195s - loss: 4.0851 - acc: 0.12 - ETA: 2187s - loss: 4.0803 - acc: 0.11 - ETA: 2168s - loss: 4.0953 - acc: 0.11 - ETA: 2148s - loss: 4.0983 - acc: 0.11 - ETA: 2131s - loss: 4.0856 - acc: 0.11 - ETA: 2126s - loss: 4.0817 - acc: 0.11 - ETA: 2111s - loss: 4.0820 - acc: 0.10 - ETA: 2106s - loss: 4.0607 - acc: 0.10 - ETA: 2098s - loss: 4.0662 - acc: 0.10 - ETA: 2105s - loss: 4.0676 - acc: 0.10 - ETA: 2093s - loss: 4.0712 - acc: 0.10 - ETA: 2081s - loss: 4.0750 - acc: 0.10 - ETA: 2073s - loss: 4.0731 - acc: 0.10 - ETA: 2058s - loss: 4.0705 - acc: 0.10 - ETA: 2045s - loss: 4.0757 - acc: 0.10 - ETA: 2033s - loss: 4.0741 - acc: 0.10 - ETA: 2019s - loss: 4.0789 - acc: 0.10 - ETA: 2002s - loss: 4.0759 - acc: 0.10 - ETA: 1989s - loss: 4.0728 - acc: 0.10 - ETA: 1974s - loss: 4.0759 - acc: 0.10 - ETA: 1960s - loss: 4.0812 - acc: 0.10 - ETA: 1949s - loss: 4.0761 - acc: 0.10 - ETA: 1941s - loss: 4.0809 - acc: 0.10 - ETA: 1928s - loss: 4.0717 - acc: 0.10 - ETA: 1919s - loss: 4.0693 - acc: 0.10 - ETA: 1906s - loss: 4.0737 - acc: 0.09 - ETA: 1894s - loss: 4.0659 - acc: 0.10 - ETA: 1886s - loss: 4.0715 - acc: 0.10 - ETA: 1873s - loss: 4.0677 - acc: 0.10 - ETA: 1864s - loss: 4.0714 - acc: 0.09 - ETA: 1854s - loss: 4.0727 - acc: 0.09 - ETA: 1852s - loss: 4.0732 - acc: 0.09 - ETA: 1847s - loss: 4.0685 - acc: 0.09 - ETA: 1843s - loss: 4.0664 - acc: 0.09 - ETA: 1834s - loss: 4.0661 - acc: 0.09 - ETA: 1823s - loss: 4.0706 - acc: 0.09 - ETA: 1816s - loss: 4.0695 - acc: 0.10 - ETA: 1805s - loss: 4.0704 - acc: 0.09 - ETA: 1794s - loss: 4.0697 - acc: 0.09 - ETA: 1787s - loss: 4.0673 - acc: 0.09 - ETA: 1782s - loss: 4.0668 - acc: 0.09 - ETA: 1776s - loss: 4.0629 - acc: 0.09 - ETA: 1767s - loss: 4.0618 - acc: 0.10 - ETA: 1757s - loss: 4.0638 - acc: 0.10 - ETA: 1749s - loss: 4.0600 - acc: 0.10 - ETA: 1737s - loss: 4.0607 - acc: 0.10 - ETA: 1725s - loss: 4.0642 - acc: 0.09 - ETA: 1717s - loss: 4.0626 - acc: 0.09 - ETA: 1709s - loss: 4.0560 - acc: 0.10 - ETA: 1702s - loss: 4.0563 - acc: 0.10 - ETA: 1694s - loss: 4.0572 - acc: 0.10 - ETA: 1687s - loss: 4.0548 - acc: 0.10 - ETA: 1679s - loss: 4.0511 - acc: 0.10 - ETA: 1670s - loss: 4.0484 - acc: 0.10 - ETA: 1664s - loss: 4.0521 - acc: 0.10 - ETA: 1657s - loss: 4.0498 - acc: 0.10 - ETA: 1647s - loss: 4.0513 - acc: 0.10 - ETA: 1640s - loss: 4.0523 - acc: 0.10 - ETA: 1630s - loss: 4.0496 - acc: 0.10 - ETA: 1620s - loss: 4.0451 - acc: 0.10 - ETA: 1610s - loss: 4.0442 - acc: 0.10 - ETA: 1600s - loss: 4.0427 - acc: 0.10 - ETA: 1589s - loss: 4.0428 - acc: 0.10 - ETA: 1579s - loss: 4.0428 - acc: 0.09 - ETA: 1568s - loss: 4.0414 - acc: 0.10 - ETA: 1558s - loss: 4.0402 - acc: 0.10 - ETA: 1548s - loss: 4.0402 - acc: 0.10 - ETA: 1538s - loss: 4.0443 - acc: 0.10 - ETA: 1527s - loss: 4.0430 - acc: 0.10 - ETA: 1516s - loss: 4.0426 - acc: 0.10 - ETA: 1507s - loss: 4.0415 - acc: 0.10 - ETA: 1497s - loss: 4.0397 - acc: 0.10 - ETA: 1488s - loss: 4.0361 - acc: 0.10 - ETA: 1479s - loss: 4.0409 - acc: 0.10 - ETA: 1468s - loss: 4.0396 - acc: 0.10 - ETA: 1457s - loss: 4.0354 - acc: 0.10 - ETA: 1446s - loss: 4.0371 - acc: 0.10 - ETA: 1435s - loss: 4.0369 - acc: 0.10 - ETA: 1425s - loss: 4.0337 - acc: 0.10 - ETA: 1416s - loss: 4.0307 - acc: 0.10 - ETA: 1406s - loss: 4.0352 - acc: 0.10 - ETA: 1397s - loss: 4.0375 - acc: 0.10 - ETA: 1388s - loss: 4.0374 - acc: 0.10 - ETA: 1378s - loss: 4.0405 - acc: 0.10 - ETA: 1370s - loss: 4.0380 - acc: 0.10 - ETA: 1361s - loss: 4.0378 - acc: 0.10 - ETA: 1353s - loss: 4.0396 - acc: 0.10 - ETA: 1344s - loss: 4.0382 - acc: 0.10 - ETA: 1336s - loss: 4.0371 - acc: 0.10 - ETA: 1327s - loss: 4.0375 - acc: 0.10 - ETA: 1317s - loss: 4.0380 - acc: 0.10 - ETA: 1308s - loss: 4.0350 - acc: 0.10 - ETA: 1300s - loss: 4.0362 - acc: 0.10 - ETA: 1292s - loss: 4.0384 - acc: 0.10 - ETA: 1284s - loss: 4.0357 - acc: 0.10 - ETA: 1276s - loss: 4.0339 - acc: 0.10 - ETA: 1268s - loss: 4.0315 - acc: 0.10 - ETA: 1259s - loss: 4.0303 - acc: 0.10 - ETA: 1252s - loss: 4.0310 - acc: 0.10 - ETA: 1246s - loss: 4.0306 - acc: 0.10 - ETA: 1237s - loss: 4.0296 - acc: 0.10 - ETA: 1230s - loss: 4.0292 - acc: 0.10 - ETA: 1222s - loss: 4.0280 - acc: 0.10 - ETA: 1215s - loss: 4.0279 - acc: 0.10 - ETA: 1206s - loss: 4.0258 - acc: 0.10 - ETA: 1198s - loss: 4.0300 - acc: 0.10 - ETA: 1191s - loss: 4.0299 - acc: 0.10 - ETA: 1184s - loss: 4.0309 - acc: 0.10 - ETA: 1185s - loss: 4.0320 - acc: 0.10 - ETA: 1187s - loss: 4.0314 - acc: 0.10 - ETA: 1187s - loss: 4.0342 - acc: 0.10 - ETA: 1186s - loss: 4.0301 - acc: 0.10 - ETA: 1187s - loss: 4.0281 - acc: 0.10 - ETA: 1185s - loss: 4.0279 - acc: 0.10 - ETA: 1183s - loss: 4.0243 - acc: 0.10 - ETA: 1179s - loss: 4.0252 - acc: 0.10 - ETA: 1175s - loss: 4.0251 - acc: 0.10 - ETA: 1170s - loss: 4.0261 - acc: 0.10 - ETA: 1165s - loss: 4.0291 - acc: 0.10 - ETA: 1161s - loss: 4.0286 - acc: 0.10 - ETA: 1157s - loss: 4.0265 - acc: 0.10 - ETA: 1154s - loss: 4.0310 - acc: 0.10 - ETA: 1148s - loss: 4.0295 - acc: 0.10 - ETA: 1142s - loss: 4.0284 - acc: 0.10 - ETA: 1138s - loss: 4.0282 - acc: 0.10 - ETA: 1132s - loss: 4.0270 - acc: 0.10 - ETA: 1126s - loss: 4.0236 - acc: 0.10 - ETA: 1120s - loss: 4.0212 - acc: 0.10 - ETA: 1117s - loss: 4.0227 - acc: 0.10 - ETA: 1110s - loss: 4.0226 - acc: 0.10 - ETA: 1104s - loss: 4.0207 - acc: 0.10 - ETA: 1097s - loss: 4.0187 - acc: 0.10 - ETA: 1089s - loss: 4.0184 - acc: 0.10 - ETA: 1082s - loss: 4.0174 - acc: 0.10 - ETA: 1074s - loss: 4.0199 - acc: 0.10 - ETA: 1067s - loss: 4.0188 - acc: 0.10 - ETA: 1060s - loss: 4.0152 - acc: 0.10 - ETA: 1053s - loss: 4.0140 - acc: 0.10 - ETA: 1046s - loss: 4.0145 - acc: 0.10 - ETA: 1040s - loss: 4.0162 - acc: 0.10 - ETA: 1033s - loss: 4.0144 - acc: 0.10 - ETA: 1026s - loss: 4.0137 - acc: 0.10 - ETA: 1018s - loss: 4.0116 - acc: 0.10 - ETA: 1011s - loss: 4.0141 - acc: 0.10 - ETA: 1004s - loss: 4.0115 - acc: 0.10 - ETA: 997s - loss: 4.0099 - acc: 0.1026 - ETA: 989s - loss: 4.0108 - acc: 0.102 - ETA: 981s - loss: 4.0097 - acc: 0.102 - ETA: 974s - loss: 4.0111 - acc: 0.102 - ETA: 968s - loss: 4.0131 - acc: 0.102 - ETA: 961s - loss: 4.0110 - acc: 0.103 - ETA: 955s - loss: 4.0102 - acc: 0.103 - ETA: 948s - loss: 4.0085 - acc: 0.103 - ETA: 941s - loss: 4.0111 - acc: 0.102 - ETA: 934s - loss: 4.0124 - acc: 0.102 - ETA: 927s - loss: 4.0140 - acc: 0.102 - ETA: 920s - loss: 4.0149 - acc: 0.102 - ETA: 912s - loss: 4.0129 - acc: 0.102 - ETA: 906s - loss: 4.0120 - acc: 0.103 - ETA: 900s - loss: 4.0116 - acc: 0.103 - ETA: 893s - loss: 4.0115 - acc: 0.103 - ETA: 886s - loss: 4.0106 - acc: 0.103 - ETA: 880s - loss: 4.0101 - acc: 0.103 - ETA: 873s - loss: 4.0103 - acc: 0.103 - ETA: 866s - loss: 4.0094 - acc: 0.103 - ETA: 860s - loss: 4.0086 - acc: 0.103 - ETA: 853s - loss: 4.0081 - acc: 0.103 - ETA: 847s - loss: 4.0077 - acc: 0.103 - ETA: 841s - loss: 4.0089 - acc: 0.103 - ETA: 836s - loss: 4.0086 - acc: 0.103 - ETA: 830s - loss: 4.0061 - acc: 0.103 - ETA: 824s - loss: 4.0056 - acc: 0.104 - ETA: 818s - loss: 4.0032 - acc: 0.104 - ETA: 812s - loss: 4.0044 - acc: 0.103 - ETA: 805s - loss: 4.0039 - acc: 0.10426660/6680 [============================>.] - ETA: 799s - loss: 4.0043 - acc: 0.104 - ETA: 792s - loss: 4.0042 - acc: 0.104 - ETA: 785s - loss: 4.0044 - acc: 0.104 - ETA: 779s - loss: 4.0047 - acc: 0.104 - ETA: 772s - loss: 4.0039 - acc: 0.104 - ETA: 766s - loss: 4.0035 - acc: 0.103 - ETA: 760s - loss: 4.0014 - acc: 0.104 - ETA: 753s - loss: 3.9987 - acc: 0.104 - ETA: 747s - loss: 4.0000 - acc: 0.104 - ETA: 740s - loss: 4.0012 - acc: 0.104 - ETA: 734s - loss: 4.0014 - acc: 0.103 - ETA: 727s - loss: 4.0022 - acc: 0.103 - ETA: 721s - loss: 4.0063 - acc: 0.103 - ETA: 714s - loss: 4.0060 - acc: 0.103 - ETA: 708s - loss: 4.0062 - acc: 0.103 - ETA: 702s - loss: 4.0059 - acc: 0.103 - ETA: 695s - loss: 4.0063 - acc: 0.102 - ETA: 688s - loss: 4.0064 - acc: 0.102 - ETA: 682s - loss: 4.0063 - acc: 0.102 - ETA: 675s - loss: 4.0061 - acc: 0.102 - ETA: 669s - loss: 4.0051 - acc: 0.102 - ETA: 663s - loss: 4.0042 - acc: 0.102 - ETA: 657s - loss: 4.0027 - acc: 0.102 - ETA: 650s - loss: 4.0027 - acc: 0.102 - ETA: 644s - loss: 4.0029 - acc: 0.102 - ETA: 637s - loss: 4.0021 - acc: 0.102 - ETA: 631s - loss: 4.0016 - acc: 0.103 - ETA: 624s - loss: 4.0012 - acc: 0.103 - ETA: 618s - loss: 4.0007 - acc: 0.103 - ETA: 612s - loss: 4.0008 - acc: 0.103 - ETA: 605s - loss: 3.9995 - acc: 0.104 - ETA: 600s - loss: 4.0002 - acc: 0.103 - ETA: 594s - loss: 4.0010 - acc: 0.104 - ETA: 588s - loss: 4.0003 - acc: 0.104 - ETA: 581s - loss: 3.9982 - acc: 0.105 - ETA: 575s - loss: 4.0000 - acc: 0.105 - ETA: 569s - loss: 3.9979 - acc: 0.106 - ETA: 563s - loss: 3.9962 - acc: 0.106 - ETA: 556s - loss: 3.9941 - acc: 0.106 - ETA: 550s - loss: 3.9948 - acc: 0.105 - ETA: 544s - loss: 3.9944 - acc: 0.105 - ETA: 538s - loss: 3.9933 - acc: 0.105 - ETA: 531s - loss: 3.9941 - acc: 0.106 - ETA: 525s - loss: 3.9935 - acc: 0.106 - ETA: 519s - loss: 3.9944 - acc: 0.106 - ETA: 513s - loss: 3.9957 - acc: 0.106 - ETA: 507s - loss: 3.9963 - acc: 0.105 - ETA: 501s - loss: 3.9959 - acc: 0.105 - ETA: 495s - loss: 3.9961 - acc: 0.105 - ETA: 489s - loss: 3.9942 - acc: 0.105 - ETA: 482s - loss: 3.9962 - acc: 0.105 - ETA: 476s - loss: 3.9955 - acc: 0.105 - ETA: 470s - loss: 3.9940 - acc: 0.105 - ETA: 464s - loss: 3.9952 - acc: 0.105 - ETA: 458s - loss: 3.9955 - acc: 0.105 - ETA: 452s - loss: 3.9954 - acc: 0.106 - ETA: 445s - loss: 3.9980 - acc: 0.106 - ETA: 439s - loss: 3.9979 - acc: 0.106 - ETA: 433s - loss: 3.9990 - acc: 0.106 - ETA: 427s - loss: 3.9984 - acc: 0.106 - ETA: 420s - loss: 3.9972 - acc: 0.106 - ETA: 414s - loss: 3.9954 - acc: 0.107 - ETA: 408s - loss: 3.9953 - acc: 0.107 - ETA: 401s - loss: 3.9961 - acc: 0.107 - ETA: 395s - loss: 3.9965 - acc: 0.107 - ETA: 389s - loss: 3.9954 - acc: 0.107 - ETA: 383s - loss: 3.9958 - acc: 0.106 - ETA: 377s - loss: 3.9964 - acc: 0.106 - ETA: 370s - loss: 3.9967 - acc: 0.106 - ETA: 364s - loss: 3.9944 - acc: 0.106 - ETA: 357s - loss: 3.9979 - acc: 0.106 - ETA: 351s - loss: 3.9969 - acc: 0.106 - ETA: 344s - loss: 3.9982 - acc: 0.105 - ETA: 338s - loss: 3.9971 - acc: 0.106 - ETA: 332s - loss: 3.9975 - acc: 0.105 - ETA: 325s - loss: 3.9987 - acc: 0.105 - ETA: 319s - loss: 3.9981 - acc: 0.106 - ETA: 313s - loss: 3.9982 - acc: 0.105 - ETA: 306s - loss: 3.9975 - acc: 0.105 - ETA: 300s - loss: 3.9978 - acc: 0.105 - ETA: 294s - loss: 3.9971 - acc: 0.105 - ETA: 288s - loss: 3.9975 - acc: 0.105 - ETA: 281s - loss: 3.9956 - acc: 0.106 - ETA: 275s - loss: 3.9955 - acc: 0.106 - ETA: 269s - loss: 3.9951 - acc: 0.106 - ETA: 263s - loss: 3.9949 - acc: 0.106 - ETA: 257s - loss: 3.9934 - acc: 0.106 - ETA: 250s - loss: 3.9922 - acc: 0.106 - ETA: 244s - loss: 3.9920 - acc: 0.106 - ETA: 238s - loss: 3.9914 - acc: 0.106 - ETA: 232s - loss: 3.9892 - acc: 0.107 - ETA: 226s - loss: 3.9881 - acc: 0.107 - ETA: 220s - loss: 3.9863 - acc: 0.107 - ETA: 213s - loss: 3.9868 - acc: 0.107 - ETA: 207s - loss: 3.9873 - acc: 0.107 - ETA: 201s - loss: 3.9866 - acc: 0.106 - ETA: 195s - loss: 3.9858 - acc: 0.107 - ETA: 189s - loss: 3.9864 - acc: 0.107 - ETA: 183s - loss: 3.9864 - acc: 0.107 - ETA: 177s - loss: 3.9858 - acc: 0.107 - ETA: 171s - loss: 3.9859 - acc: 0.107 - ETA: 165s - loss: 3.9858 - acc: 0.107 - ETA: 159s - loss: 3.9853 - acc: 0.107 - ETA: 153s - loss: 3.9865 - acc: 0.107 - ETA: 147s - loss: 3.9867 - acc: 0.107 - ETA: 141s - loss: 3.9876 - acc: 0.106 - ETA: 135s - loss: 3.9885 - acc: 0.107 - ETA: 129s - loss: 3.9872 - acc: 0.107 - ETA: 123s - loss: 3.9884 - acc: 0.107 - ETA: 117s - loss: 3.9868 - acc: 0.107 - ETA: 111s - loss: 3.9861 - acc: 0.107 - ETA: 105s - loss: 3.9849 - acc: 0.107 - ETA: 99s - loss: 3.9841 - acc: 0.108 - ETA: 93s - loss: 3.9828 - acc: 0.10 - ETA: 87s - loss: 3.9834 - acc: 0.10 - ETA: 81s - loss: 3.9831 - acc: 0.10 - ETA: 75s - loss: 3.9842 - acc: 0.10 - ETA: 70s - loss: 3.9839 - acc: 0.10 - ETA: 64s - loss: 3.9825 - acc: 0.10 - ETA: 58s - loss: 3.9829 - acc: 0.10 - ETA: 52s - loss: 3.9839 - acc: 0.10 - ETA: 46s - loss: 3.9844 - acc: 0.10 - ETA: 40s - loss: 3.9832 - acc: 0.10 - ETA: 34s - loss: 3.9830 - acc: 0.10 - ETA: 29s - loss: 3.9836 - acc: 0.10 - ETA: 23s - loss: 3.9841 - acc: 0.10 - ETA: 17s - loss: 3.9863 - acc: 0.10 - ETA: 11s - loss: 3.9873 - acc: 0.10 - ETA: 5s - loss: 3.9859 - acc: 0.1072 Epoch 00002: val_loss improved from 4.29654 to 4.16885, saving model to saved_models/weights.best.from_scratch.hdf5 6680/6680 [==============================] - 1959s - loss: 3.9842 - acc: 0.1070 - val_loss: 4.1688 - val_acc: 0.0886 Epoch 4/5 4080/6680 [=================>............] - ETA: 1979s - loss: 3.7862 - acc: 0.05 - ETA: 1826s - loss: 3.9114 - acc: 0.07 - ETA: 1679s - loss: 3.8184 - acc: 0.11 - ETA: 1535s - loss: 3.5581 - acc: 0.17 - ETA: 1429s - loss: 3.4988 - acc: 0.18 - ETA: 1422s - loss: 3.4573 - acc: 0.17 - ETA: 1397s - loss: 3.4653 - acc: 0.17 - ETA: 1383s - loss: 3.4831 - acc: 0.17 - ETA: 1382s - loss: 3.5007 - acc: 0.17 - ETA: 1379s - loss: 3.4771 - acc: 0.17 - ETA: 1391s - loss: 3.4667 - acc: 0.17 - ETA: 1391s - loss: 3.4649 - acc: 0.17 - ETA: 1386s - loss: 3.4455 - acc: 0.18 - ETA: 1377s - loss: 3.4343 - acc: 0.17 - ETA: 1391s - loss: 3.4350 - acc: 0.18 - ETA: 1392s - loss: 3.4748 - acc: 0.18 - ETA: 1394s - loss: 3.4948 - acc: 0.18 - ETA: 1397s - loss: 3.4890 - acc: 0.18 - ETA: 1411s - loss: 3.4969 - acc: 0.18 - ETA: 1405s - loss: 3.5015 - acc: 0.18 - ETA: 1408s - loss: 3.4968 - acc: 0.18 - ETA: 1402s - loss: 3.4910 - acc: 0.18 - ETA: 1391s - loss: 3.4864 - acc: 0.18 - ETA: 1391s - loss: 3.4936 - acc: 0.17 - ETA: 1391s - loss: 3.4764 - acc: 0.18 - ETA: 1387s - loss: 3.4613 - acc: 0.18 - ETA: 1387s - loss: 3.4592 - acc: 0.18 - ETA: 1380s - loss: 3.4510 - acc: 0.19 - ETA: 1379s - loss: 3.4366 - acc: 0.19 - ETA: 1386s - loss: 3.4555 - acc: 0.18 - ETA: 1379s - loss: 3.4695 - acc: 0.18 - ETA: 1371s - loss: 3.4597 - acc: 0.18 - ETA: 1380s - loss: 3.4565 - acc: 0.18 - ETA: 1369s - loss: 3.4488 - acc: 0.18 - ETA: 1367s - loss: 3.4297 - acc: 0.18 - ETA: 1362s - loss: 3.4349 - acc: 0.18 - ETA: 1356s - loss: 3.4432 - acc: 0.18 - ETA: 1353s - loss: 3.4481 - acc: 0.18 - ETA: 1347s - loss: 3.4504 - acc: 0.18 - ETA: 1345s - loss: 3.4481 - acc: 0.19 - ETA: 1333s - loss: 3.4374 - acc: 0.19 - ETA: 1330s - loss: 3.4286 - acc: 0.19 - ETA: 1326s - loss: 3.4202 - acc: 0.20 - ETA: 1321s - loss: 3.4063 - acc: 0.20 - ETA: 1319s - loss: 3.4194 - acc: 0.20 - ETA: 1311s - loss: 3.4151 - acc: 0.20 - ETA: 1302s - loss: 3.4191 - acc: 0.20 - ETA: 1293s - loss: 3.4185 - acc: 0.20 - ETA: 1285s - loss: 3.4193 - acc: 0.20 - ETA: 1279s - loss: 3.4189 - acc: 0.20 - ETA: 1274s - loss: 3.4246 - acc: 0.20 - ETA: 1265s - loss: 3.4267 - acc: 0.20 - ETA: 1258s - loss: 3.4239 - acc: 0.20 - ETA: 1252s - loss: 3.4191 - acc: 0.20 - ETA: 1247s - loss: 3.4269 - acc: 0.20 - ETA: 1243s - loss: 3.4305 - acc: 0.20 - ETA: 1237s - loss: 3.4299 - acc: 0.20 - ETA: 1227s - loss: 3.4213 - acc: 0.20 - ETA: 1221s - loss: 3.4236 - acc: 0.20 - ETA: 1215s - loss: 3.4230 - acc: 0.20 - ETA: 1209s - loss: 3.4307 - acc: 0.20 - ETA: 1201s - loss: 3.4267 - acc: 0.20 - ETA: 1198s - loss: 3.4196 - acc: 0.20 - ETA: 1195s - loss: 3.4155 - acc: 0.20 - ETA: 1189s - loss: 3.4115 - acc: 0.20 - ETA: 1183s - loss: 3.4129 - acc: 0.20 - ETA: 1176s - loss: 3.4121 - acc: 0.20 - ETA: 1170s - loss: 3.4107 - acc: 0.20 - ETA: 1165s - loss: 3.4036 - acc: 0.20 - ETA: 1158s - loss: 3.4032 - acc: 0.20 - ETA: 1152s - loss: 3.4031 - acc: 0.20 - ETA: 1149s - loss: 3.3945 - acc: 0.20 - ETA: 1140s - loss: 3.3902 - acc: 0.20 - ETA: 1134s - loss: 3.3858 - acc: 0.20 - ETA: 1127s - loss: 3.3880 - acc: 0.20 - ETA: 1119s - loss: 3.3840 - acc: 0.20 - ETA: 1115s - loss: 3.3845 - acc: 0.20 - ETA: 1116s - loss: 3.3860 - acc: 0.20 - ETA: 1113s - loss: 3.3892 - acc: 0.20 - ETA: 1109s - loss: 3.3897 - acc: 0.20 - ETA: 1104s - loss: 3.3885 - acc: 0.20 - ETA: 1099s - loss: 3.3892 - acc: 0.20 - ETA: 1097s - loss: 3.3955 - acc: 0.20 - ETA: 1093s - loss: 3.3958 - acc: 0.20 - ETA: 1090s - loss: 3.3974 - acc: 0.20 - ETA: 1084s - loss: 3.3953 - acc: 0.20 - ETA: 1079s - loss: 3.3956 - acc: 0.20 - ETA: 1070s - loss: 3.3998 - acc: 0.20 - ETA: 1065s - loss: 3.4039 - acc: 0.20 - ETA: 1060s - loss: 3.4057 - acc: 0.20 - ETA: 1055s - loss: 3.4120 - acc: 0.20 - ETA: 1051s - loss: 3.4098 - acc: 0.20 - ETA: 1046s - loss: 3.4079 - acc: 0.20 - ETA: 1040s - loss: 3.4049 - acc: 0.20 - ETA: 1037s - loss: 3.4046 - acc: 0.20 - ETA: 1034s - loss: 3.4065 - acc: 0.20 - ETA: 1031s - loss: 3.4076 - acc: 0.20 - ETA: 1028s - loss: 3.4038 - acc: 0.20 - ETA: 1024s - loss: 3.4024 - acc: 0.20 - ETA: 1020s - loss: 3.4028 - acc: 0.20 - ETA: 1016s - loss: 3.4077 - acc: 0.20 - ETA: 1012s - loss: 3.4100 - acc: 0.20 - ETA: 1008s - loss: 3.4115 - acc: 0.20 - ETA: 1004s - loss: 3.4109 - acc: 0.20 - ETA: 1002s - loss: 3.4035 - acc: 0.20 - ETA: 996s - loss: 3.4031 - acc: 0.2042 - ETA: 992s - loss: 3.3961 - acc: 0.205 - ETA: 986s - loss: 3.3956 - acc: 0.206 - ETA: 981s - loss: 3.3973 - acc: 0.205 - ETA: 977s - loss: 3.3959 - acc: 0.205 - ETA: 973s - loss: 3.3948 - acc: 0.204 - ETA: 966s - loss: 3.3920 - acc: 0.204 - ETA: 962s - loss: 3.3865 - acc: 0.206 - ETA: 956s - loss: 3.3875 - acc: 0.206 - ETA: 951s - loss: 3.3960 - acc: 0.205 - ETA: 947s - loss: 3.3956 - acc: 0.205 - ETA: 942s - loss: 3.3983 - acc: 0.205 - ETA: 934s - loss: 3.3974 - acc: 0.205 - ETA: 930s - loss: 3.3971 - acc: 0.204 - ETA: 926s - loss: 3.3974 - acc: 0.204 - ETA: 922s - loss: 3.3984 - acc: 0.204 - ETA: 917s - loss: 3.3941 - acc: 0.204 - ETA: 912s - loss: 3.3896 - acc: 0.206 - ETA: 908s - loss: 3.3824 - acc: 0.208 - ETA: 905s - loss: 3.3829 - acc: 0.208 - ETA: 900s - loss: 3.3791 - acc: 0.208 - ETA: 897s - loss: 3.3814 - acc: 0.208 - ETA: 893s - loss: 3.3800 - acc: 0.208 - ETA: 889s - loss: 3.3806 - acc: 0.208 - ETA: 885s - loss: 3.3853 - acc: 0.206 - ETA: 880s - loss: 3.3862 - acc: 0.206 - ETA: 877s - loss: 3.3828 - acc: 0.206 - ETA: 873s - loss: 3.3775 - acc: 0.208 - ETA: 869s - loss: 3.3756 - acc: 0.209 - ETA: 864s - loss: 3.3789 - acc: 0.208 - ETA: 860s - loss: 3.3775 - acc: 0.208 - ETA: 856s - loss: 3.3792 - acc: 0.208 - ETA: 851s - loss: 3.3813 - acc: 0.207 - ETA: 847s - loss: 3.3820 - acc: 0.207 - ETA: 842s - loss: 3.3833 - acc: 0.206 - ETA: 837s - loss: 3.3831 - acc: 0.207 - ETA: 832s - loss: 3.3828 - acc: 0.207 - ETA: 827s - loss: 3.3828 - acc: 0.206 - ETA: 824s - loss: 3.3830 - acc: 0.206 - ETA: 819s - loss: 3.3869 - acc: 0.206 - ETA: 815s - loss: 3.3838 - acc: 0.207 - ETA: 810s - loss: 3.3827 - acc: 0.207 - ETA: 806s - loss: 3.3809 - acc: 0.207 - ETA: 802s - loss: 3.3817 - acc: 0.207 - ETA: 798s - loss: 3.3801 - acc: 0.208 - ETA: 794s - loss: 3.3818 - acc: 0.208 - ETA: 789s - loss: 3.3824 - acc: 0.209 - ETA: 784s - loss: 3.3772 - acc: 0.211 - ETA: 781s - loss: 3.3749 - acc: 0.210 - ETA: 777s - loss: 3.3739 - acc: 0.210 - ETA: 773s - loss: 3.3757 - acc: 0.209 - ETA: 768s - loss: 3.3746 - acc: 0.209 - ETA: 764s - loss: 3.3742 - acc: 0.208 - ETA: 758s - loss: 3.3750 - acc: 0.208 - ETA: 755s - loss: 3.3721 - acc: 0.208 - ETA: 759s - loss: 3.3714 - acc: 0.209 - ETA: 7914s - loss: 3.3717 - acc: 0.20 - ETA: 7837s - loss: 3.3710 - acc: 0.20 - ETA: 7752s - loss: 3.3702 - acc: 0.20 - ETA: 7673s - loss: 3.3720 - acc: 0.20 - ETA: 7592s - loss: 3.3733 - acc: 0.20 - ETA: 7511s - loss: 3.3757 - acc: 0.20 - ETA: 7432s - loss: 3.3758 - acc: 0.20 - ETA: 7352s - loss: 3.3756 - acc: 0.20 - ETA: 7273s - loss: 3.3719 - acc: 0.21 - ETA: 7193s - loss: 3.3714 - acc: 0.21 - ETA: 7117s - loss: 3.3695 - acc: 0.21 - ETA: 7041s - loss: 3.3673 - acc: 0.21 - ETA: 6968s - loss: 3.3726 - acc: 0.21 - ETA: 6895s - loss: 3.3700 - acc: 0.21 - ETA: 6820s - loss: 3.3711 - acc: 0.21 - ETA: 6745s - loss: 3.3677 - acc: 0.21 - ETA: 6672s - loss: 3.3670 - acc: 0.21 - ETA: 6602s - loss: 3.3674 - acc: 0.21 - ETA: 6533s - loss: 3.3652 - acc: 0.21 - ETA: 6464s - loss: 3.3646 - acc: 0.21 - ETA: 6394s - loss: 3.3662 - acc: 0.21 - ETA: 6325s - loss: 3.3687 - acc: 0.21 - ETA: 6257s - loss: 3.3721 - acc: 0.21 - ETA: 6190s - loss: 3.3715 - acc: 0.21 - ETA: 6121s - loss: 3.3712 - acc: 0.21 - ETA: 6053s - loss: 3.3690 - acc: 0.21 - ETA: 5986s - loss: 3.3686 - acc: 0.21 - ETA: 5920s - loss: 3.3727 - acc: 0.21 - ETA: 5854s - loss: 3.3731 - acc: 0.21 - ETA: 5788s - loss: 3.3728 - acc: 0.21 - ETA: 5723s - loss: 3.3710 - acc: 0.21 - ETA: 5659s - loss: 3.3732 - acc: 0.21 - ETA: 5596s - loss: 3.3711 - acc: 0.21 - ETA: 5533s - loss: 3.3723 - acc: 0.21 - ETA: 5472s - loss: 3.3741 - acc: 0.21 - ETA: 5411s - loss: 3.3728 - acc: 0.21 - ETA: 5350s - loss: 3.3740 - acc: 0.21 - ETA: 5290s - loss: 3.3737 - acc: 0.21 - ETA: 5229s - loss: 3.3737 - acc: 0.21 - ETA: 5168s - loss: 3.3742 - acc: 0.21 - ETA: 5109s - loss: 3.3736 - acc: 0.21 - ETA: 5050s - loss: 3.3735 - acc: 0.21 - ETA: 4992s - loss: 3.3725 - acc: 0.21256660/6680 [============================>.] - ETA: 4933s - loss: 3.3718 - acc: 0.21 - ETA: 4875s - loss: 3.3713 - acc: 0.21 - ETA: 4817s - loss: 3.3713 - acc: 0.21 - ETA: 4761s - loss: 3.3733 - acc: 0.21 - ETA: 4704s - loss: 3.3734 - acc: 0.21 - ETA: 4648s - loss: 3.3761 - acc: 0.21 - ETA: 4592s - loss: 3.3734 - acc: 0.21 - ETA: 4537s - loss: 3.3722 - acc: 0.21 - ETA: 4482s - loss: 3.3710 - acc: 0.21 - ETA: 4428s - loss: 3.3733 - acc: 0.21 - ETA: 4374s - loss: 3.3779 - acc: 0.21 - ETA: 4320s - loss: 3.3775 - acc: 0.21 - ETA: 4268s - loss: 3.3761 - acc: 0.21 - ETA: 4215s - loss: 3.3754 - acc: 0.21 - ETA: 4163s - loss: 3.3776 - acc: 0.21 - ETA: 4111s - loss: 3.3808 - acc: 0.21 - ETA: 4061s - loss: 3.3814 - acc: 0.21 - ETA: 4011s - loss: 3.3813 - acc: 0.21 - ETA: 3960s - loss: 3.3825 - acc: 0.20 - ETA: 3910s - loss: 3.3830 - acc: 0.21 - ETA: 3860s - loss: 3.3837 - acc: 0.21 - ETA: 3811s - loss: 3.3800 - acc: 0.21 - ETA: 3761s - loss: 3.3777 - acc: 0.21 - ETA: 3712s - loss: 3.3767 - acc: 0.21 - ETA: 3664s - loss: 3.3771 - acc: 0.21 - ETA: 3616s - loss: 3.3763 - acc: 0.21 - ETA: 3568s - loss: 3.3751 - acc: 0.21 - ETA: 3521s - loss: 3.3755 - acc: 0.21 - ETA: 3474s - loss: 3.3743 - acc: 0.21 - ETA: 3427s - loss: 3.3750 - acc: 0.21 - ETA: 3382s - loss: 3.3739 - acc: 0.21 - ETA: 3337s - loss: 3.3752 - acc: 0.21 - ETA: 3291s - loss: 3.3743 - acc: 0.21 - ETA: 3246s - loss: 3.3730 - acc: 0.21 - ETA: 3201s - loss: 3.3734 - acc: 0.21 - ETA: 3157s - loss: 3.3737 - acc: 0.21 - ETA: 3113s - loss: 3.3756 - acc: 0.21 - ETA: 3069s - loss: 3.3757 - acc: 0.21 - ETA: 3025s - loss: 3.3763 - acc: 0.21 - ETA: 2981s - loss: 3.3757 - acc: 0.21 - ETA: 2938s - loss: 3.3750 - acc: 0.21 - ETA: 2895s - loss: 3.3741 - acc: 0.21 - ETA: 2853s - loss: 3.3748 - acc: 0.21 - ETA: 2811s - loss: 3.3733 - acc: 0.21 - ETA: 2769s - loss: 3.3740 - acc: 0.21 - ETA: 2728s - loss: 3.3727 - acc: 0.21 - ETA: 2687s - loss: 3.3711 - acc: 0.21 - ETA: 2646s - loss: 3.3681 - acc: 0.21 - ETA: 2606s - loss: 3.3682 - acc: 0.21 - ETA: 2565s - loss: 3.3686 - acc: 0.21 - ETA: 2526s - loss: 3.3700 - acc: 0.21 - ETA: 2487s - loss: 3.3702 - acc: 0.21 - ETA: 2448s - loss: 3.3684 - acc: 0.21 - ETA: 2410s - loss: 3.3670 - acc: 0.21 - ETA: 2371s - loss: 3.3689 - acc: 0.21 - ETA: 2332s - loss: 3.3717 - acc: 0.21 - ETA: 2294s - loss: 3.3724 - acc: 0.21 - ETA: 2256s - loss: 3.3722 - acc: 0.21 - ETA: 2218s - loss: 3.3703 - acc: 0.21 - ETA: 2181s - loss: 3.3695 - acc: 0.21 - ETA: 2143s - loss: 3.3688 - acc: 0.21 - ETA: 2106s - loss: 3.3696 - acc: 0.21 - ETA: 2069s - loss: 3.3698 - acc: 0.21 - ETA: 2033s - loss: 3.3698 - acc: 0.21 - ETA: 1996s - loss: 3.3680 - acc: 0.21 - ETA: 1960s - loss: 3.3671 - acc: 0.21 - ETA: 1924s - loss: 3.3661 - acc: 0.21 - ETA: 1888s - loss: 3.3685 - acc: 0.21 - ETA: 1852s - loss: 3.3678 - acc: 0.21 - ETA: 1817s - loss: 3.3674 - acc: 0.21 - ETA: 1782s - loss: 3.3672 - acc: 0.21 - ETA: 1746s - loss: 3.3659 - acc: 0.21 - ETA: 1711s - loss: 3.3664 - acc: 0.21 - ETA: 1676s - loss: 3.3645 - acc: 0.21 - ETA: 1642s - loss: 3.3653 - acc: 0.21 - ETA: 1608s - loss: 3.3660 - acc: 0.21 - ETA: 1573s - loss: 3.3638 - acc: 0.21 - ETA: 1540s - loss: 3.3655 - acc: 0.21 - ETA: 1506s - loss: 3.3643 - acc: 0.21 - ETA: 1473s - loss: 3.3671 - acc: 0.21 - ETA: 1440s - loss: 3.3668 - acc: 0.21 - ETA: 1407s - loss: 3.3657 - acc: 0.21 - ETA: 1374s - loss: 3.3666 - acc: 0.21 - ETA: 1341s - loss: 3.3651 - acc: 0.21 - ETA: 1308s - loss: 3.3664 - acc: 0.21 - ETA: 1276s - loss: 3.3659 - acc: 0.21 - ETA: 1245s - loss: 3.3663 - acc: 0.21 - ETA: 1213s - loss: 3.3660 - acc: 0.21 - ETA: 1181s - loss: 3.3638 - acc: 0.21 - ETA: 1150s - loss: 3.3641 - acc: 0.21 - ETA: 1118s - loss: 3.3643 - acc: 0.21 - ETA: 1086s - loss: 3.3632 - acc: 0.21 - ETA: 1055s - loss: 3.3639 - acc: 0.21 - ETA: 1024s - loss: 3.3638 - acc: 0.21 - ETA: 993s - loss: 3.3628 - acc: 0.2156 - ETA: 962s - loss: 3.3647 - acc: 0.215 - ETA: 932s - loss: 3.3657 - acc: 0.215 - ETA: 901s - loss: 3.3650 - acc: 0.215 - ETA: 871s - loss: 3.3659 - acc: 0.215 - ETA: 841s - loss: 3.3657 - acc: 0.215 - ETA: 811s - loss: 3.3651 - acc: 0.215 - ETA: 781s - loss: 3.3631 - acc: 0.215 - ETA: 751s - loss: 3.3650 - acc: 0.215 - ETA: 721s - loss: 3.3641 - acc: 0.215 - ETA: 692s - loss: 3.3647 - acc: 0.215 - ETA: 663s - loss: 3.3641 - acc: 0.215 - ETA: 633s - loss: 3.3643 - acc: 0.215 - ETA: 604s - loss: 3.3648 - acc: 0.215 - ETA: 575s - loss: 3.3651 - acc: 0.215 - ETA: 547s - loss: 3.3646 - acc: 0.215 - ETA: 518s - loss: 3.3639 - acc: 0.215 - ETA: 489s - loss: 3.3649 - acc: 0.214 - ETA: 461s - loss: 3.3655 - acc: 0.214 - ETA: 433s - loss: 3.3658 - acc: 0.214 - ETA: 405s - loss: 3.3659 - acc: 0.214 - ETA: 377s - loss: 3.3658 - acc: 0.214 - ETA: 349s - loss: 3.3663 - acc: 0.214 - ETA: 321s - loss: 3.3657 - acc: 0.214 - ETA: 294s - loss: 3.3656 - acc: 0.214 - ETA: 266s - loss: 3.3650 - acc: 0.214 - ETA: 239s - loss: 3.3671 - acc: 0.214 - ETA: 212s - loss: 3.3657 - acc: 0.213 - ETA: 185s - loss: 3.3668 - acc: 0.213 - ETA: 158s - loss: 3.3680 - acc: 0.213 - ETA: 131s - loss: 3.3683 - acc: 0.212 - ETA: 105s - loss: 3.3684 - acc: 0.212 - ETA: 78s - loss: 3.3683 - acc: 0.212 - ETA: 52s - loss: 3.3682 - acc: 0.21 - ETA: 26s - loss: 3.3672 - acc: 0.2122Epoch 00003: val_loss improved from 4.16885 to 4.13443, saving model to saved_models/weights.best.from_scratch.hdf5 6680/6680 [==============================] - 8754s - loss: 3.3687 - acc: 0.2124 - val_loss: 4.1344 - val_acc: 0.0874 Epoch 5/5 4080/6680 [=================>............] - ETA: 2399s - loss: 2.7436 - acc: 0.30 - ETA: 2476s - loss: 2.5123 - acc: 0.35 - ETA: 2500s - loss: 2.5977 - acc: 0.36 - ETA: 2494s - loss: 2.4750 - acc: 0.38 - ETA: 2428s - loss: 2.5514 - acc: 0.38 - ETA: 2394s - loss: 2.6987 - acc: 0.36 - ETA: 2391s - loss: 2.8057 - acc: 0.35 - ETA: 2395s - loss: 2.8616 - acc: 0.34 - ETA: 2373s - loss: 2.8814 - acc: 0.33 - ETA: 2381s - loss: 2.8533 - acc: 0.33 - ETA: 2402s - loss: 2.8281 - acc: 0.33 - ETA: 2398s - loss: 2.8175 - acc: 0.34 - ETA: 2397s - loss: 2.7805 - acc: 0.35 - ETA: 2382s - loss: 2.7593 - acc: 0.35 - ETA: 2342s - loss: 2.7344 - acc: 0.35 - ETA: 2314s - loss: 2.7181 - acc: 0.35 - ETA: 2322s - loss: 2.6691 - acc: 0.36 - ETA: 2305s - loss: 2.6628 - acc: 0.35 - ETA: 2291s - loss: 2.6902 - acc: 0.35 - ETA: 2280s - loss: 2.6630 - acc: 0.35 - ETA: 2261s - loss: 2.6798 - acc: 0.35 - ETA: 2269s - loss: 2.7012 - acc: 0.35 - ETA: 2248s - loss: 2.6879 - acc: 0.35 - ETA: 2230s - loss: 2.6630 - acc: 0.35 - ETA: 2262s - loss: 2.6654 - acc: 0.36 - ETA: 2259s - loss: 2.6515 - acc: 0.36 - ETA: 2241s - loss: 2.6197 - acc: 0.37 - ETA: 2235s - loss: 2.6255 - acc: 0.36 - ETA: 2252s - loss: 2.6049 - acc: 0.37 - ETA: 2242s - loss: 2.6074 - acc: 0.37 - ETA: 2235s - loss: 2.5936 - acc: 0.37 - ETA: 2232s - loss: 2.5952 - acc: 0.37 - ETA: 2221s - loss: 2.5773 - acc: 0.37 - ETA: 2215s - loss: 2.5829 - acc: 0.37 - ETA: 2210s - loss: 2.5848 - acc: 0.37 - ETA: 2203s - loss: 2.5807 - acc: 0.37 - ETA: 2196s - loss: 2.5808 - acc: 0.37 - ETA: 2189s - loss: 2.5778 - acc: 0.37 - ETA: 2183s - loss: 2.5584 - acc: 0.38 - ETA: 2171s - loss: 2.5581 - acc: 0.38 - ETA: 2156s - loss: 2.5536 - acc: 0.38 - ETA: 2144s - loss: 2.5557 - acc: 0.38 - ETA: 2133s - loss: 2.5514 - acc: 0.38 - ETA: 2123s - loss: 2.5453 - acc: 0.38 - ETA: 2104s - loss: 2.5445 - acc: 0.38 - ETA: 2089s - loss: 2.5500 - acc: 0.38 - ETA: 2085s - loss: 2.5354 - acc: 0.38 - ETA: 2072s - loss: 2.5415 - acc: 0.38 - ETA: 2065s - loss: 2.5369 - acc: 0.38 - ETA: 2059s - loss: 2.5241 - acc: 0.38 - ETA: 2049s - loss: 2.5259 - acc: 0.38 - ETA: 2049s - loss: 2.5206 - acc: 0.38 - ETA: 2036s - loss: 2.5129 - acc: 0.38 - ETA: 2028s - loss: 2.5193 - acc: 0.38 - ETA: 2023s - loss: 2.5150 - acc: 0.38 - ETA: 2013s - loss: 2.5145 - acc: 0.38 - ETA: 1999s - loss: 2.5099 - acc: 0.38 - ETA: 1984s - loss: 2.5125 - acc: 0.38 - ETA: 1978s - loss: 2.5304 - acc: 0.38 - ETA: 1965s - loss: 2.5107 - acc: 0.39 - ETA: 1954s - loss: 2.5037 - acc: 0.39 - ETA: 1944s - loss: 2.4944 - acc: 0.39 - ETA: 1936s - loss: 2.4888 - acc: 0.39 - ETA: 1928s - loss: 2.4895 - acc: 0.39 - ETA: 1918s - loss: 2.4853 - acc: 0.39 - ETA: 1908s - loss: 2.4807 - acc: 0.40 - ETA: 1897s - loss: 2.4913 - acc: 0.39 - ETA: 1890s - loss: 2.4930 - acc: 0.39 - ETA: 1880s - loss: 2.4892 - acc: 0.39 - ETA: 1877s - loss: 2.5067 - acc: 0.39 - ETA: 1876s - loss: 2.5069 - acc: 0.39 - ETA: 1872s - loss: 2.5135 - acc: 0.39 - ETA: 1861s - loss: 2.5142 - acc: 0.39 - ETA: 1851s - loss: 2.5096 - acc: 0.39 - ETA: 1839s - loss: 2.5095 - acc: 0.39 - ETA: 1830s - loss: 2.5170 - acc: 0.39 - ETA: 1825s - loss: 2.5139 - acc: 0.39 - ETA: 1815s - loss: 2.5098 - acc: 0.39 - ETA: 1810s - loss: 2.5082 - acc: 0.39 - ETA: 1816s - loss: 2.5018 - acc: 0.40 - ETA: 1814s - loss: 2.4999 - acc: 0.40 - ETA: 1809s - loss: 2.4987 - acc: 0.40 - ETA: 1801s - loss: 2.5010 - acc: 0.40 - ETA: 1791s - loss: 2.5017 - acc: 0.40 - ETA: 1785s - loss: 2.5036 - acc: 0.40 - ETA: 1776s - loss: 2.5023 - acc: 0.40 - ETA: 1767s - loss: 2.5033 - acc: 0.40 - ETA: 1755s - loss: 2.5132 - acc: 0.39 - ETA: 1746s - loss: 2.5265 - acc: 0.39 - ETA: 1740s - loss: 2.5321 - acc: 0.39 - ETA: 1733s - loss: 2.5299 - acc: 0.39 - ETA: 1727s - loss: 2.5293 - acc: 0.39 - ETA: 1721s - loss: 2.5275 - acc: 0.40 - ETA: 1714s - loss: 2.5231 - acc: 0.40 - ETA: 1709s - loss: 2.5280 - acc: 0.39 - ETA: 1702s - loss: 2.5301 - acc: 0.39 - ETA: 1692s - loss: 2.5306 - acc: 0.39 - ETA: 1686s - loss: 2.5266 - acc: 0.39 - ETA: 1683s - loss: 2.5307 - acc: 0.39 - ETA: 1674s - loss: 2.5297 - acc: 0.39 - ETA: 1665s - loss: 2.5259 - acc: 0.39 - ETA: 1656s - loss: 2.5215 - acc: 0.40 - ETA: 1647s - loss: 2.5209 - acc: 0.39 - ETA: 1644s - loss: 2.5181 - acc: 0.39 - ETA: 1637s - loss: 2.5155 - acc: 0.39 - ETA: 1627s - loss: 2.5104 - acc: 0.40 - ETA: 1619s - loss: 2.5067 - acc: 0.40 - ETA: 1610s - loss: 2.5032 - acc: 0.40 - ETA: 1599s - loss: 2.5024 - acc: 0.40 - ETA: 1593s - loss: 2.4929 - acc: 0.40 - ETA: 1585s - loss: 2.4958 - acc: 0.40 - ETA: 1575s - loss: 2.4982 - acc: 0.40 - ETA: 1565s - loss: 2.4970 - acc: 0.40 - ETA: 1557s - loss: 2.4994 - acc: 0.40 - ETA: 1549s - loss: 2.5014 - acc: 0.40 - ETA: 1540s - loss: 2.5068 - acc: 0.40 - ETA: 1532s - loss: 2.5117 - acc: 0.39 - ETA: 1524s - loss: 2.5120 - acc: 0.39 - ETA: 1517s - loss: 2.5096 - acc: 0.39 - ETA: 1509s - loss: 2.5116 - acc: 0.39 - ETA: 1500s - loss: 2.5084 - acc: 0.39 - ETA: 1494s - loss: 2.5088 - acc: 0.39 - ETA: 1485s - loss: 2.5052 - acc: 0.39 - ETA: 1475s - loss: 2.5027 - acc: 0.40 - ETA: 1466s - loss: 2.5021 - acc: 0.40 - ETA: 1456s - loss: 2.4997 - acc: 0.40 - ETA: 1448s - loss: 2.5026 - acc: 0.39 - ETA: 1440s - loss: 2.5037 - acc: 0.39 - ETA: 1431s - loss: 2.5090 - acc: 0.39 - ETA: 1421s - loss: 2.5124 - acc: 0.39 - ETA: 1413s - loss: 2.5096 - acc: 0.39 - ETA: 1403s - loss: 2.5085 - acc: 0.39 - ETA: 1393s - loss: 2.5117 - acc: 0.39 - ETA: 1384s - loss: 2.5164 - acc: 0.39 - ETA: 1375s - loss: 2.5125 - acc: 0.39 - ETA: 1366s - loss: 2.5140 - acc: 0.39 - ETA: 1357s - loss: 2.5167 - acc: 0.39 - ETA: 1349s - loss: 2.5187 - acc: 0.39 - ETA: 1340s - loss: 2.5211 - acc: 0.39 - ETA: 1331s - loss: 2.5199 - acc: 0.39 - ETA: 1323s - loss: 2.5198 - acc: 0.39 - ETA: 1315s - loss: 2.5182 - acc: 0.39 - ETA: 1307s - loss: 2.5182 - acc: 0.39 - ETA: 1298s - loss: 2.5134 - acc: 0.39 - ETA: 1289s - loss: 2.5130 - acc: 0.39 - ETA: 1281s - loss: 2.5107 - acc: 0.39 - ETA: 1273s - loss: 2.5045 - acc: 0.39 - ETA: 1265s - loss: 2.5077 - acc: 0.39 - ETA: 1257s - loss: 2.5101 - acc: 0.39 - ETA: 1248s - loss: 2.5107 - acc: 0.39 - ETA: 1241s - loss: 2.5138 - acc: 0.39 - ETA: 1234s - loss: 2.5141 - acc: 0.39 - ETA: 1226s - loss: 2.5198 - acc: 0.39 - ETA: 1217s - loss: 2.5191 - acc: 0.39 - ETA: 1209s - loss: 2.5153 - acc: 0.39 - ETA: 1202s - loss: 2.5127 - acc: 0.39 - ETA: 1195s - loss: 2.5117 - acc: 0.39 - ETA: 1187s - loss: 2.5077 - acc: 0.40 - ETA: 1178s - loss: 2.5100 - acc: 0.40 - ETA: 1170s - loss: 2.5086 - acc: 0.39 - ETA: 1164s - loss: 2.5068 - acc: 0.39 - ETA: 1161s - loss: 2.5046 - acc: 0.40 - ETA: 1158s - loss: 2.5005 - acc: 0.40 - ETA: 1152s - loss: 2.5017 - acc: 0.40 - ETA: 1145s - loss: 2.4977 - acc: 0.40 - ETA: 1138s - loss: 2.4979 - acc: 0.40 - ETA: 1132s - loss: 2.4993 - acc: 0.40 - ETA: 1126s - loss: 2.4996 - acc: 0.40 - ETA: 1121s - loss: 2.4994 - acc: 0.40 - ETA: 1113s - loss: 2.4980 - acc: 0.40 - ETA: 1106s - loss: 2.4999 - acc: 0.40 - ETA: 1099s - loss: 2.5004 - acc: 0.40 - ETA: 1093s - loss: 2.4959 - acc: 0.40 - ETA: 1086s - loss: 2.4932 - acc: 0.40 - ETA: 1079s - loss: 2.4966 - acc: 0.40 - ETA: 1072s - loss: 2.4934 - acc: 0.40 - ETA: 1063s - loss: 2.4995 - acc: 0.39 - ETA: 1056s - loss: 2.4988 - acc: 0.39 - ETA: 1050s - loss: 2.4977 - acc: 0.39 - ETA: 1042s - loss: 2.4976 - acc: 0.39 - ETA: 1035s - loss: 2.4967 - acc: 0.39 - ETA: 1029s - loss: 2.4986 - acc: 0.39 - ETA: 1022s - loss: 2.4968 - acc: 0.39 - ETA: 1014s - loss: 2.4976 - acc: 0.39 - ETA: 1007s - loss: 2.4937 - acc: 0.39 - ETA: 1000s - loss: 2.4978 - acc: 0.39 - ETA: 993s - loss: 2.5019 - acc: 0.3979 - ETA: 986s - loss: 2.5028 - acc: 0.397 - ETA: 978s - loss: 2.5005 - acc: 0.397 - ETA: 971s - loss: 2.5015 - acc: 0.396 - ETA: 964s - loss: 2.4995 - acc: 0.397 - ETA: 956s - loss: 2.5020 - acc: 0.396 - ETA: 949s - loss: 2.5020 - acc: 0.396 - ETA: 941s - loss: 2.4992 - acc: 0.396 - ETA: 934s - loss: 2.4984 - acc: 0.397 - ETA: 926s - loss: 2.4959 - acc: 0.398 - ETA: 919s - loss: 2.4993 - acc: 0.397 - ETA: 912s - loss: 2.5019 - acc: 0.397 - ETA: 904s - loss: 2.5028 - acc: 0.396 - ETA: 897s - loss: 2.5051 - acc: 0.396 - ETA: 890s - loss: 2.5067 - acc: 0.395 - ETA: 883s - loss: 2.5052 - acc: 0.396 - ETA: 876s - loss: 2.5038 - acc: 0.396 - ETA: 869s - loss: 2.5052 - acc: 0.39516660/6680 [============================>.] - ETA: 862s - loss: 2.5038 - acc: 0.395 - ETA: 855s - loss: 2.5015 - acc: 0.396 - ETA: 848s - loss: 2.5014 - acc: 0.395 - ETA: 841s - loss: 2.5038 - acc: 0.395 - ETA: 836s - loss: 2.5041 - acc: 0.395 - ETA: 830s - loss: 2.5071 - acc: 0.394 - ETA: 824s - loss: 2.5068 - acc: 0.394 - ETA: 817s - loss: 2.5056 - acc: 0.394 - ETA: 810s - loss: 2.5019 - acc: 0.395 - ETA: 804s - loss: 2.5028 - acc: 0.395 - ETA: 797s - loss: 2.5008 - acc: 0.395 - ETA: 790s - loss: 2.4989 - acc: 0.395 - ETA: 783s - loss: 2.4967 - acc: 0.395 - ETA: 776s - loss: 2.4944 - acc: 0.396 - ETA: 771s - loss: 2.4979 - acc: 0.395 - ETA: 764s - loss: 2.4958 - acc: 0.396 - ETA: 758s - loss: 2.4955 - acc: 0.396 - ETA: 751s - loss: 2.4940 - acc: 0.397 - ETA: 744s - loss: 2.4932 - acc: 0.396 - ETA: 737s - loss: 2.4920 - acc: 0.396 - ETA: 730s - loss: 2.4915 - acc: 0.396 - ETA: 723s - loss: 2.4885 - acc: 0.397 - ETA: 717s - loss: 2.4880 - acc: 0.397 - ETA: 710s - loss: 2.4880 - acc: 0.397 - ETA: 702s - loss: 2.4875 - acc: 0.397 - ETA: 695s - loss: 2.4874 - acc: 0.397 - ETA: 688s - loss: 2.4897 - acc: 0.396 - ETA: 682s - loss: 2.4911 - acc: 0.396 - ETA: 675s - loss: 2.4900 - acc: 0.395 - ETA: 668s - loss: 2.4906 - acc: 0.396 - ETA: 661s - loss: 2.4886 - acc: 0.396 - ETA: 655s - loss: 2.4874 - acc: 0.396 - ETA: 648s - loss: 2.4855 - acc: 0.397 - ETA: 641s - loss: 2.4861 - acc: 0.396 - ETA: 634s - loss: 2.4873 - acc: 0.397 - ETA: 627s - loss: 2.4891 - acc: 0.396 - ETA: 620s - loss: 2.4886 - acc: 0.397 - ETA: 612s - loss: 2.4907 - acc: 0.397 - ETA: 605s - loss: 2.4904 - acc: 0.397 - ETA: 598s - loss: 2.4871 - acc: 0.397 - ETA: 592s - loss: 2.4867 - acc: 0.398 - ETA: 585s - loss: 2.4902 - acc: 0.397 - ETA: 578s - loss: 2.4899 - acc: 0.397 - ETA: 570s - loss: 2.4884 - acc: 0.397 - ETA: 563s - loss: 2.4877 - acc: 0.397 - ETA: 556s - loss: 2.4878 - acc: 0.397 - ETA: 550s - loss: 2.4881 - acc: 0.397 - ETA: 543s - loss: 2.4871 - acc: 0.397 - ETA: 536s - loss: 2.4855 - acc: 0.397 - ETA: 529s - loss: 2.4858 - acc: 0.397 - ETA: 523s - loss: 2.4861 - acc: 0.397 - ETA: 516s - loss: 2.4865 - acc: 0.397 - ETA: 509s - loss: 2.4882 - acc: 0.397 - ETA: 502s - loss: 2.4888 - acc: 0.397 - ETA: 495s - loss: 2.4888 - acc: 0.397 - ETA: 489s - loss: 2.4885 - acc: 0.397 - ETA: 482s - loss: 2.4869 - acc: 0.397 - ETA: 475s - loss: 2.4861 - acc: 0.397 - ETA: 468s - loss: 2.4854 - acc: 0.397 - ETA: 461s - loss: 2.4825 - acc: 0.397 - ETA: 454s - loss: 2.4813 - acc: 0.397 - ETA: 447s - loss: 2.4799 - acc: 0.397 - ETA: 440s - loss: 2.4826 - acc: 0.396 - ETA: 434s - loss: 2.4851 - acc: 0.396 - ETA: 427s - loss: 2.4851 - acc: 0.396 - ETA: 420s - loss: 2.4887 - acc: 0.396 - ETA: 413s - loss: 2.4896 - acc: 0.395 - ETA: 406s - loss: 2.4910 - acc: 0.395 - ETA: 399s - loss: 2.4912 - acc: 0.395 - ETA: 392s - loss: 2.4902 - acc: 0.395 - ETA: 385s - loss: 2.4934 - acc: 0.395 - ETA: 379s - loss: 2.4915 - acc: 0.395 - ETA: 372s - loss: 2.4926 - acc: 0.396 - ETA: 365s - loss: 2.4910 - acc: 0.396 - ETA: 358s - loss: 2.4879 - acc: 0.397 - ETA: 351s - loss: 2.4868 - acc: 0.397 - ETA: 344s - loss: 2.4872 - acc: 0.397 - ETA: 338s - loss: 2.4890 - acc: 0.396 - ETA: 332s - loss: 2.4926 - acc: 0.395 - ETA: 325s - loss: 2.4947 - acc: 0.395 - ETA: 319s - loss: 2.4973 - acc: 0.394 - ETA: 312s - loss: 2.4977 - acc: 0.394 - ETA: 306s - loss: 2.4986 - acc: 0.394 - ETA: 300s - loss: 2.4992 - acc: 0.394 - ETA: 293s - loss: 2.4989 - acc: 0.394 - ETA: 287s - loss: 2.4971 - acc: 0.394 - ETA: 280s - loss: 2.4994 - acc: 0.394 - ETA: 273s - loss: 2.4986 - acc: 0.394 - ETA: 267s - loss: 2.4993 - acc: 0.394 - ETA: 260s - loss: 2.5007 - acc: 0.394 - ETA: 254s - loss: 2.5004 - acc: 0.394 - ETA: 247s - loss: 2.5013 - acc: 0.394 - ETA: 241s - loss: 2.4992 - acc: 0.395 - ETA: 235s - loss: 2.5017 - acc: 0.395 - ETA: 228s - loss: 2.4998 - acc: 0.395 - ETA: 221s - loss: 2.5019 - acc: 0.394 - ETA: 215s - loss: 2.5015 - acc: 0.394 - ETA: 208s - loss: 2.5007 - acc: 0.395 - ETA: 202s - loss: 2.5006 - acc: 0.395 - ETA: 196s - loss: 2.4996 - acc: 0.395 - ETA: 190s - loss: 2.5005 - acc: 0.395 - ETA: 183s - loss: 2.5021 - acc: 0.395 - ETA: 177s - loss: 2.5019 - acc: 0.394 - ETA: 170s - loss: 2.5028 - acc: 0.394 - ETA: 164s - loss: 2.5019 - acc: 0.395 - ETA: 157s - loss: 2.5012 - acc: 0.395 - ETA: 151s - loss: 2.5020 - acc: 0.395 - ETA: 144s - loss: 2.5035 - acc: 0.394 - ETA: 138s - loss: 2.5012 - acc: 0.395 - ETA: 131s - loss: 2.5001 - acc: 0.396 - ETA: 125s - loss: 2.4988 - acc: 0.396 - ETA: 118s - loss: 2.4974 - acc: 0.396 - ETA: 112s - loss: 2.4974 - acc: 0.396 - ETA: 105s - loss: 2.4973 - acc: 0.396 - ETA: 99s - loss: 2.4995 - acc: 0.395 - ETA: 92s - loss: 2.4982 - acc: 0.39 - ETA: 85s - loss: 2.4964 - acc: 0.39 - ETA: 79s - loss: 2.4958 - acc: 0.39 - ETA: 72s - loss: 2.4962 - acc: 0.39 - ETA: 66s - loss: 2.4967 - acc: 0.39 - ETA: 59s - loss: 2.4958 - acc: 0.39 - ETA: 52s - loss: 2.4952 - acc: 0.39 - ETA: 46s - loss: 2.4933 - acc: 0.39 - ETA: 39s - loss: 2.4933 - acc: 0.39 - ETA: 33s - loss: 2.4941 - acc: 0.39 - ETA: 26s - loss: 2.4942 - acc: 0.39 - ETA: 19s - loss: 2.4940 - acc: 0.39 - ETA: 13s - loss: 2.4934 - acc: 0.39 - ETA: 6s - loss: 2.4952 - acc: 0.3967 Epoch 00004: val_loss did not improve 6680/6680 [==============================] - 2238s - loss: 2.4946 - acc: 0.3970 - val_loss: 4.4974 - val_acc: 0.0910
Out[15]:
<keras.callbacks.History at 0x23327dc0898>
In [16]:
model.load_weights('saved_models/weights.best.from_scratch.hdf5')
Test the Model¶
Try out your model on the test dataset of dog images. Ensure that your test accuracy is greater than 1%.
In [17]:
# get index of predicted dog breed for each image in test set
dog_breed_predictions = [np.argmax(model.predict(np.expand_dims(tensor, axis=0))) for tensor in test_tensors]
# report test accuracy
test_accuracy = 100*np.sum(np.array(dog_breed_predictions)==np.argmax(test_targets, axis=1))/len(dog_breed_predictions)
print('Test accuracy: %.4f%%' % test_accuracy)
Test accuracy: 10.7656%
In [18]:
bottleneck_features = np.load('bottleneck_features/DogVGG16Data.npz')
train_VGG16 = bottleneck_features['train']
valid_VGG16 = bottleneck_features['valid']
test_VGG16 = bottleneck_features['test']
Model Architecture¶
The model uses the the pre-trained VGG-16 model as a fixed feature extractor, where the last convolutional output of VGG-16 is fed as input to our model. We only add a global average pooling layer and a fully connected layer, where the latter contains one node for each dog category and is equipped with a softmax.
In [19]:
VGG16_model = Sequential()
VGG16_model.add(GlobalAveragePooling2D(input_shape=train_VGG16.shape[1:]))
VGG16_model.add(Dense(133, activation='softmax'))
VGG16_model.summary()
_________________________________________________________________ Layer (type) Output Shape Param # ================================================================= global_average_pooling2d_1 ( (None, 512) 0 _________________________________________________________________ dense_3 (Dense) (None, 133) 68229 ================================================================= Total params: 68,229.0 Trainable params: 68,229.0 Non-trainable params: 0.0 _________________________________________________________________
Compile the Model¶
In [20]:
VGG16_model.compile(loss='categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])
Train the Model¶
In [21]:
checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.VGG16.hdf5',
verbose=1, save_best_only=True)
VGG16_model.fit(train_VGG16, train_targets,
validation_data=(valid_VGG16, valid_targets),
epochs=20, batch_size=20, callbacks=[checkpointer], verbose=1)
Train on 6680 samples, validate on 835 samples Epoch 1/20 6640/6680 [============================>.] - ETA: 389s - loss: 15.0877 - acc: 0.0000e+ - ETA: 139s - loss: 15.0899 - acc: 0.0333 - ETA: 111s - loss: 15.1993 - acc: 0.02 - ETA: 92s - loss: 15.1926 - acc: 0.0300 - ETA: 68s - loss: 15.3235 - acc: 0.021 - ETA: 54s - loss: 15.2596 - acc: 0.022 - ETA: 46s - loss: 15.1185 - acc: 0.027 - ETA: 40s - loss: 15.2161 - acc: 0.023 - ETA: 35s - loss: 15.1675 - acc: 0.026 - ETA: 34s - loss: 15.0882 - acc: 0.028 - ETA: 34s - loss: 15.1488 - acc: 0.026 - ETA: 31s - loss: 15.0810 - acc: 0.026 - ETA: 32s - loss: 15.0207 - acc: 0.025 - ETA: 32s - loss: 14.9636 - acc: 0.026 - ETA: 31s - loss: 14.9757 - acc: 0.025 - ETA: 29s - loss: 14.8939 - acc: 0.025 - ETA: 28s - loss: 14.9047 - acc: 0.024 - ETA: 28s - loss: 14.9109 - acc: 0.023 - ETA: 28s - loss: 14.8959 - acc: 0.022 - ETA: 28s - loss: 14.9052 - acc: 0.021 - ETA: 27s - loss: 14.8536 - acc: 0.022 - ETA: 27s - loss: 14.8134 - acc: 0.023 - ETA: 27s - loss: 14.7911 - acc: 0.024 - ETA: 26s - loss: 14.7409 - acc: 0.023 - ETA: 26s - loss: 14.7681 - acc: 0.022 - ETA: 26s - loss: 14.7692 - acc: 0.022 - ETA: 25s - loss: 14.7529 - acc: 0.021 - ETA: 24s - loss: 14.6326 - acc: 0.026 - ETA: 23s - loss: 14.6522 - acc: 0.026 - ETA: 22s - loss: 14.6228 - acc: 0.028 - ETA: 21s - loss: 14.5306 - acc: 0.030 - ETA: 20s - loss: 14.5048 - acc: 0.030 - ETA: 19s - loss: 14.4795 - acc: 0.031 - ETA: 18s - loss: 14.4216 - acc: 0.035 - ETA: 18s - loss: 14.3842 - acc: 0.037 - ETA: 17s - loss: 14.3451 - acc: 0.038 - ETA: 17s - loss: 14.2250 - acc: 0.040 - ETA: 16s - loss: 14.2189 - acc: 0.040 - ETA: 15s - loss: 14.2404 - acc: 0.039 - ETA: 15s - loss: 14.2385 - acc: 0.041 - ETA: 14s - loss: 14.1819 - acc: 0.042 - ETA: 14s - loss: 14.1368 - acc: 0.043 - ETA: 13s - loss: 14.1085 - acc: 0.044 - ETA: 13s - loss: 14.1081 - acc: 0.044 - ETA: 12s - loss: 14.0810 - acc: 0.045 - ETA: 12s - loss: 14.0737 - acc: 0.045 - ETA: 12s - loss: 14.0801 - acc: 0.045 - ETA: 11s - loss: 14.0513 - acc: 0.046 - ETA: 11s - loss: 14.0376 - acc: 0.048 - ETA: 11s - loss: 14.0373 - acc: 0.048 - ETA: 11s - loss: 14.0029 - acc: 0.049 - ETA: 10s - loss: 13.9668 - acc: 0.050 - ETA: 10s - loss: 13.9527 - acc: 0.051 - ETA: 10s - loss: 13.9194 - acc: 0.053 - ETA: 10s - loss: 13.9137 - acc: 0.053 - ETA: 10s - loss: 13.8978 - acc: 0.054 - ETA: 9s - loss: 13.8492 - acc: 0.056 - ETA: 9s - loss: 13.8201 - acc: 0.05 - ETA: 9s - loss: 13.8171 - acc: 0.05 - ETA: 9s - loss: 13.7708 - acc: 0.05 - ETA: 9s - loss: 13.7644 - acc: 0.05 - ETA: 9s - loss: 13.7785 - acc: 0.05 - ETA: 9s - loss: 13.7431 - acc: 0.06 - ETA: 9s - loss: 13.7291 - acc: 0.06 - ETA: 9s - loss: 13.7319 - acc: 0.06 - ETA: 9s - loss: 13.7252 - acc: 0.06 - ETA: 9s - loss: 13.7307 - acc: 0.06 - ETA: 8s - loss: 13.7405 - acc: 0.06 - ETA: 8s - loss: 13.7213 - acc: 0.06 - ETA: 8s - loss: 13.6940 - acc: 0.06 - ETA: 8s - loss: 13.6607 - acc: 0.06 - ETA: 8s - loss: 13.6447 - acc: 0.06 - ETA: 8s - loss: 13.6458 - acc: 0.06 - ETA: 8s - loss: 13.6269 - acc: 0.06 - ETA: 7s - loss: 13.6240 - acc: 0.06 - ETA: 7s - loss: 13.6170 - acc: 0.06 - ETA: 7s - loss: 13.5944 - acc: 0.06 - ETA: 7s - loss: 13.5773 - acc: 0.06 - ETA: 7s - loss: 13.5631 - acc: 0.06 - ETA: 7s - loss: 13.5551 - acc: 0.06 - ETA: 7s - loss: 13.5251 - acc: 0.06 - ETA: 7s - loss: 13.5249 - acc: 0.06 - ETA: 6s - loss: 13.4964 - acc: 0.06 - ETA: 6s - loss: 13.4876 - acc: 0.06 - ETA: 6s - loss: 13.4511 - acc: 0.07 - ETA: 6s - loss: 13.4297 - acc: 0.07 - ETA: 6s - loss: 13.4090 - acc: 0.07 - ETA: 6s - loss: 13.3914 - acc: 0.07 - ETA: 6s - loss: 13.3818 - acc: 0.07 - ETA: 6s - loss: 13.3576 - acc: 0.07 - ETA: 6s - loss: 13.3336 - acc: 0.07 - ETA: 5s - loss: 13.3095 - acc: 0.07 - ETA: 5s - loss: 13.2941 - acc: 0.08 - ETA: 5s - loss: 13.2822 - acc: 0.08 - ETA: 5s - loss: 13.2876 - acc: 0.08 - ETA: 5s - loss: 13.2693 - acc: 0.08 - ETA: 5s - loss: 13.2699 - acc: 0.08 - ETA: 5s - loss: 13.2679 - acc: 0.08 - ETA: 5s - loss: 13.2697 - acc: 0.08 - ETA: 5s - loss: 13.2654 - acc: 0.08 - ETA: 5s - loss: 13.2709 - acc: 0.08 - ETA: 5s - loss: 13.2670 - acc: 0.08 - ETA: 5s - loss: 13.2606 - acc: 0.08 - ETA: 5s - loss: 13.2568 - acc: 0.08 - ETA: 5s - loss: 13.2366 - acc: 0.08 - ETA: 5s - loss: 13.2207 - acc: 0.08 - ETA: 5s - loss: 13.1986 - acc: 0.08 - ETA: 5s - loss: 13.1897 - acc: 0.08 - ETA: 5s - loss: 13.1674 - acc: 0.08 - ETA: 5s - loss: 13.1666 - acc: 0.08 - ETA: 4s - loss: 13.1527 - acc: 0.08 - ETA: 4s - loss: 13.1385 - acc: 0.08 - ETA: 4s - loss: 13.1195 - acc: 0.08 - ETA: 4s - loss: 13.1144 - acc: 0.09 - ETA: 4s - loss: 13.1168 - acc: 0.09 - ETA: 4s - loss: 13.1178 - acc: 0.09 - ETA: 4s - loss: 13.1091 - acc: 0.09 - ETA: 4s - loss: 13.1083 - acc: 0.09 - ETA: 3s - loss: 13.0894 - acc: 0.09 - ETA: 3s - loss: 13.0652 - acc: 0.09 - ETA: 3s - loss: 13.0433 - acc: 0.09 - ETA: 3s - loss: 13.0296 - acc: 0.09 - ETA: 3s - loss: 13.0287 - acc: 0.09 - ETA: 3s - loss: 13.0055 - acc: 0.09 - ETA: 3s - loss: 13.0015 - acc: 0.09 - ETA: 3s - loss: 13.0008 - acc: 0.09 - ETA: 3s - loss: 13.0069 - acc: 0.09 - ETA: 3s - loss: 13.0015 - acc: 0.09 - ETA: 3s - loss: 12.9907 - acc: 0.09 - ETA: 3s - loss: 12.9719 - acc: 0.09 - ETA: 2s - loss: 12.9511 - acc: 0.10 - ETA: 2s - loss: 12.9470 - acc: 0.10 - ETA: 2s - loss: 12.9278 - acc: 0.10 - ETA: 2s - loss: 12.9193 - acc: 0.10 - ETA: 2s - loss: 12.9127 - acc: 0.10 - ETA: 2s - loss: 12.8898 - acc: 0.10 - ETA: 2s - loss: 12.8703 - acc: 0.10 - ETA: 2s - loss: 12.8618 - acc: 0.10 - ETA: 2s - loss: 12.8389 - acc: 0.10 - ETA: 1s - loss: 12.8403 - acc: 0.10 - ETA: 1s - loss: 12.8344 - acc: 0.10 - ETA: 1s - loss: 12.8185 - acc: 0.10 - ETA: 1s - loss: 12.8115 - acc: 0.11 - ETA: 1s - loss: 12.8099 - acc: 0.11 - ETA: 1s - loss: 12.8004 - acc: 0.11 - ETA: 1s - loss: 12.7841 - acc: 0.11 - ETA: 1s - loss: 12.7750 - acc: 0.11 - ETA: 1s - loss: 12.7560 - acc: 0.11 - ETA: 1s - loss: 12.7472 - acc: 0.11 - ETA: 0s - loss: 12.7456 - acc: 0.11 - ETA: 0s - loss: 12.7428 - acc: 0.11 - ETA: 0s - loss: 12.7267 - acc: 0.11 - ETA: 0s - loss: 12.7257 - acc: 0.11 - ETA: 0s - loss: 12.7279 - acc: 0.11 - ETA: 0s - loss: 12.7130 - acc: 0.11 - ETA: 0s - loss: 12.7157 - acc: 0.11 - ETA: 0s - loss: 12.7117 - acc: 0.11 - ETA: 0s - loss: 12.7057 - acc: 0.11 - ETA: 0s - loss: 12.7014 - acc: 0.11 - ETA: 0s - loss: 12.6955 - acc: 0.11 - ETA: 0s - loss: 12.6951 - acc: 0.11 - ETA: 0s - loss: 12.6957 - acc: 0.11 - ETA: 0s - loss: 12.6960 - acc: 0.11 - ETA: 0s - loss: 12.6976 - acc: 0.11 - ETA: 0s - loss: 12.6886 - acc: 0.11 - ETA: 0s - loss: 12.6846 - acc: 0.11 - ETA: 0s - loss: 12.6683 - acc: 0.1196Epoch 00000: val_loss improved from inf to 11.29383, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 13s - loss: 12.6674 - acc: 0.1196 - val_loss: 11.2938 - val_acc: 0.2168 Epoch 2/20 6660/6680 [============================>.] - ETA: 11s - loss: 14.0524 - acc: 0.050 - ETA: 8s - loss: 12.1321 - acc: 0.175 - ETA: 7s - loss: 11.1878 - acc: 0.23 - ETA: 7s - loss: 11.4272 - acc: 0.22 - ETA: 6s - loss: 11.2179 - acc: 0.23 - ETA: 6s - loss: 11.1258 - acc: 0.24 - ETA: 6s - loss: 11.1222 - acc: 0.23 - ETA: 6s - loss: 11.1933 - acc: 0.23 - ETA: 6s - loss: 11.2466 - acc: 0.23 - ETA: 6s - loss: 11.1890 - acc: 0.22 - ETA: 6s - loss: 11.3118 - acc: 0.22 - ETA: 6s - loss: 11.2468 - acc: 0.22 - ETA: 6s - loss: 11.1597 - acc: 0.22 - ETA: 6s - loss: 11.1346 - acc: 0.22 - ETA: 6s - loss: 11.2039 - acc: 0.22 - ETA: 6s - loss: 11.2865 - acc: 0.22 - ETA: 6s - loss: 11.0935 - acc: 0.23 - ETA: 6s - loss: 11.0722 - acc: 0.23 - ETA: 6s - loss: 11.0161 - acc: 0.23 - ETA: 6s - loss: 11.0027 - acc: 0.23 - ETA: 6s - loss: 10.9195 - acc: 0.24 - ETA: 6s - loss: 10.8486 - acc: 0.24 - ETA: 6s - loss: 10.8202 - acc: 0.24 - ETA: 6s - loss: 10.8208 - acc: 0.24 - ETA: 6s - loss: 10.8183 - acc: 0.24 - ETA: 6s - loss: 10.8301 - acc: 0.24 - ETA: 6s - loss: 10.8096 - acc: 0.24 - ETA: 7s - loss: 10.7949 - acc: 0.24 - ETA: 7s - loss: 10.8142 - acc: 0.24 - ETA: 7s - loss: 10.8464 - acc: 0.24 - ETA: 7s - loss: 10.8323 - acc: 0.24 - ETA: 7s - loss: 10.8619 - acc: 0.24 - ETA: 7s - loss: 10.8289 - acc: 0.24 - ETA: 7s - loss: 10.8140 - acc: 0.24 - ETA: 7s - loss: 10.8774 - acc: 0.24 - ETA: 7s - loss: 10.9337 - acc: 0.24 - ETA: 7s - loss: 10.9579 - acc: 0.24 - ETA: 7s - loss: 10.9812 - acc: 0.23 - ETA: 6s - loss: 11.0107 - acc: 0.23 - ETA: 6s - loss: 10.9085 - acc: 0.24 - ETA: 6s - loss: 10.8747 - acc: 0.24 - ETA: 6s - loss: 10.9088 - acc: 0.24 - ETA: 6s - loss: 10.8842 - acc: 0.24 - ETA: 6s - loss: 10.8808 - acc: 0.24 - ETA: 6s - loss: 10.8402 - acc: 0.24 - ETA: 6s - loss: 10.8708 - acc: 0.24 - ETA: 6s - loss: 10.8808 - acc: 0.24 - ETA: 5s - loss: 10.8993 - acc: 0.24 - ETA: 5s - loss: 10.8859 - acc: 0.24 - ETA: 5s - loss: 10.8664 - acc: 0.24 - ETA: 5s - loss: 10.8755 - acc: 0.24 - ETA: 5s - loss: 10.8519 - acc: 0.24 - ETA: 5s - loss: 10.8415 - acc: 0.24 - ETA: 5s - loss: 10.8639 - acc: 0.24 - ETA: 5s - loss: 10.8675 - acc: 0.24 - ETA: 5s - loss: 10.8614 - acc: 0.24 - ETA: 5s - loss: 10.8603 - acc: 0.24 - ETA: 5s - loss: 10.8525 - acc: 0.24 - ETA: 5s - loss: 10.8616 - acc: 0.24 - ETA: 5s - loss: 10.8776 - acc: 0.24 - ETA: 5s - loss: 10.8791 - acc: 0.24 - ETA: 5s - loss: 10.8861 - acc: 0.24 - ETA: 5s - loss: 10.9098 - acc: 0.24 - ETA: 5s - loss: 10.9137 - acc: 0.24 - ETA: 5s - loss: 10.9234 - acc: 0.24 - ETA: 5s - loss: 10.9472 - acc: 0.24 - ETA: 5s - loss: 10.9751 - acc: 0.24 - ETA: 5s - loss: 10.9940 - acc: 0.24 - ETA: 4s - loss: 10.9735 - acc: 0.24 - ETA: 4s - loss: 10.9744 - acc: 0.24 - ETA: 4s - loss: 10.9615 - acc: 0.24 - ETA: 4s - loss: 10.9533 - acc: 0.24 - ETA: 4s - loss: 10.9519 - acc: 0.24 - ETA: 4s - loss: 10.9423 - acc: 0.24 - ETA: 4s - loss: 10.9328 - acc: 0.24 - ETA: 4s - loss: 10.9501 - acc: 0.24 - ETA: 4s - loss: 10.9385 - acc: 0.24 - ETA: 4s - loss: 10.9092 - acc: 0.24 - ETA: 4s - loss: 10.9094 - acc: 0.24 - ETA: 4s - loss: 10.9062 - acc: 0.24 - ETA: 4s - loss: 10.9077 - acc: 0.24 - ETA: 4s - loss: 10.9058 - acc: 0.24 - ETA: 4s - loss: 10.9079 - acc: 0.24 - ETA: 4s - loss: 10.9125 - acc: 0.24 - ETA: 3s - loss: 10.9139 - acc: 0.24 - ETA: 3s - loss: 10.8953 - acc: 0.24 - ETA: 3s - loss: 10.8817 - acc: 0.24 - ETA: 3s - loss: 10.8753 - acc: 0.24 - ETA: 3s - loss: 10.8752 - acc: 0.24 - ETA: 3s - loss: 10.8836 - acc: 0.24 - ETA: 3s - loss: 10.8872 - acc: 0.24 - ETA: 3s - loss: 10.8864 - acc: 0.24 - ETA: 3s - loss: 10.8953 - acc: 0.24 - ETA: 3s - loss: 10.9054 - acc: 0.24 - ETA: 3s - loss: 10.9118 - acc: 0.24 - ETA: 3s - loss: 10.9178 - acc: 0.24 - ETA: 3s - loss: 10.9227 - acc: 0.24 - ETA: 3s - loss: 10.9159 - acc: 0.24 - ETA: 3s - loss: 10.9183 - acc: 0.24 - ETA: 3s - loss: 10.9391 - acc: 0.24 - ETA: 3s - loss: 10.9374 - acc: 0.24 - ETA: 3s - loss: 10.9440 - acc: 0.24 - ETA: 3s - loss: 10.9498 - acc: 0.24 - ETA: 3s - loss: 10.9565 - acc: 0.24 - ETA: 2s - loss: 10.9614 - acc: 0.24 - ETA: 2s - loss: 10.9568 - acc: 0.24 - ETA: 2s - loss: 10.9436 - acc: 0.24 - ETA: 2s - loss: 10.9328 - acc: 0.24 - ETA: 2s - loss: 10.9263 - acc: 0.24 - ETA: 2s - loss: 10.9233 - acc: 0.24 - ETA: 2s - loss: 10.9241 - acc: 0.24 - ETA: 2s - loss: 10.9287 - acc: 0.24 - ETA: 2s - loss: 10.9311 - acc: 0.24 - ETA: 2s - loss: 10.9141 - acc: 0.24 - ETA: 2s - loss: 10.9142 - acc: 0.24 - ETA: 2s - loss: 10.9055 - acc: 0.24 - ETA: 2s - loss: 10.8943 - acc: 0.24 - ETA: 2s - loss: 10.8934 - acc: 0.24 - ETA: 2s - loss: 10.8848 - acc: 0.24 - ETA: 2s - loss: 10.8765 - acc: 0.24 - ETA: 2s - loss: 10.8618 - acc: 0.24 - ETA: 2s - loss: 10.8688 - acc: 0.24 - ETA: 2s - loss: 10.8685 - acc: 0.24 - ETA: 2s - loss: 10.8572 - acc: 0.24 - ETA: 2s - loss: 10.8633 - acc: 0.24 - ETA: 1s - loss: 10.8560 - acc: 0.24 - ETA: 1s - loss: 10.8575 - acc: 0.24 - ETA: 1s - loss: 10.8673 - acc: 0.24 - ETA: 1s - loss: 10.8610 - acc: 0.24 - ETA: 1s - loss: 10.8758 - acc: 0.24 - ETA: 1s - loss: 10.8693 - acc: 0.24 - ETA: 1s - loss: 10.8532 - acc: 0.24 - ETA: 1s - loss: 10.8543 - acc: 0.24 - ETA: 1s - loss: 10.8619 - acc: 0.24 - ETA: 1s - loss: 10.8563 - acc: 0.24 - ETA: 1s - loss: 10.8443 - acc: 0.24 - ETA: 1s - loss: 10.8584 - acc: 0.24 - ETA: 1s - loss: 10.8505 - acc: 0.24 - ETA: 1s - loss: 10.8477 - acc: 0.24 - ETA: 1s - loss: 10.8434 - acc: 0.25 - ETA: 1s - loss: 10.8521 - acc: 0.24 - ETA: 1s - loss: 10.8484 - acc: 0.24 - ETA: 0s - loss: 10.8500 - acc: 0.24 - ETA: 0s - loss: 10.8689 - acc: 0.24 - ETA: 0s - loss: 10.8632 - acc: 0.24 - ETA: 0s - loss: 10.8689 - acc: 0.24 - ETA: 0s - loss: 10.8684 - acc: 0.24 - ETA: 0s - loss: 10.8664 - acc: 0.24 - ETA: 0s - loss: 10.8617 - acc: 0.24 - ETA: 0s - loss: 10.8805 - acc: 0.24 - ETA: 0s - loss: 10.8687 - acc: 0.24 - ETA: 0s - loss: 10.8771 - acc: 0.24 - ETA: 0s - loss: 10.8759 - acc: 0.24 - ETA: 0s - loss: 10.8737 - acc: 0.24 - ETA: 0s - loss: 10.8633 - acc: 0.24 - ETA: 0s - loss: 10.8588 - acc: 0.25 - ETA: 0s - loss: 10.8543 - acc: 0.2502Epoch 00001: val_loss improved from 11.29383 to 10.56121, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 11s - loss: 10.8524 - acc: 0.2503 - val_loss: 10.5612 - val_acc: 0.2647 Epoch 3/20 6660/6680 [============================>.] - ETA: 10s - loss: 9.2108 - acc: 0.25 - ETA: 9s - loss: 10.5360 - acc: 0.23 - ETA: 9s - loss: 11.5815 - acc: 0.21 - ETA: 9s - loss: 11.2014 - acc: 0.22 - ETA: 12s - loss: 11.1117 - acc: 0.243 - ETA: 13s - loss: 11.0485 - acc: 0.250 - ETA: 14s - loss: 11.0929 - acc: 0.245 - ETA: 15s - loss: 10.8550 - acc: 0.254 - ETA: 15s - loss: 10.9016 - acc: 0.250 - ETA: 15s - loss: 10.9398 - acc: 0.250 - ETA: 15s - loss: 10.9495 - acc: 0.253 - ETA: 15s - loss: 10.9410 - acc: 0.262 - ETA: 15s - loss: 10.9150 - acc: 0.267 - ETA: 14s - loss: 10.8365 - acc: 0.276 - ETA: 14s - loss: 10.6832 - acc: 0.285 - ETA: 13s - loss: 10.7591 - acc: 0.279 - ETA: 12s - loss: 10.8349 - acc: 0.274 - ETA: 11s - loss: 10.7756 - acc: 0.273 - ETA: 10s - loss: 10.8012 - acc: 0.271 - ETA: 10s - loss: 10.8542 - acc: 0.269 - ETA: 9s - loss: 10.7578 - acc: 0.275 - ETA: 9s - loss: 10.7915 - acc: 0.27 - ETA: 9s - loss: 10.7542 - acc: 0.27 - ETA: 8s - loss: 10.7174 - acc: 0.27 - ETA: 8s - loss: 10.6976 - acc: 0.27 - ETA: 8s - loss: 10.6958 - acc: 0.27 - ETA: 8s - loss: 10.6784 - acc: 0.28 - ETA: 8s - loss: 10.7451 - acc: 0.27 - ETA: 7s - loss: 10.7290 - acc: 0.27 - ETA: 7s - loss: 10.7099 - acc: 0.27 - ETA: 7s - loss: 10.7081 - acc: 0.27 - ETA: 7s - loss: 10.6201 - acc: 0.28 - ETA: 7s - loss: 10.6684 - acc: 0.27 - ETA: 7s - loss: 10.6583 - acc: 0.27 - ETA: 7s - loss: 10.6406 - acc: 0.28 - ETA: 7s - loss: 10.5975 - acc: 0.28 - ETA: 7s - loss: 10.5752 - acc: 0.28 - ETA: 7s - loss: 10.5582 - acc: 0.28 - ETA: 7s - loss: 10.5793 - acc: 0.28 - ETA: 7s - loss: 10.5459 - acc: 0.28 - ETA: 6s - loss: 10.4951 - acc: 0.28 - ETA: 6s - loss: 10.4878 - acc: 0.29 - ETA: 6s - loss: 10.4721 - acc: 0.29 - ETA: 6s - loss: 10.4816 - acc: 0.29 - ETA: 6s - loss: 10.4192 - acc: 0.29 - ETA: 6s - loss: 10.3973 - acc: 0.29 - ETA: 6s - loss: 10.3928 - acc: 0.29 - ETA: 6s - loss: 10.3771 - acc: 0.29 - ETA: 6s - loss: 10.3262 - acc: 0.30 - ETA: 5s - loss: 10.3612 - acc: 0.30 - ETA: 5s - loss: 10.3594 - acc: 0.30 - ETA: 5s - loss: 10.3460 - acc: 0.30 - ETA: 5s - loss: 10.3299 - acc: 0.30 - ETA: 5s - loss: 10.3561 - acc: 0.30 - ETA: 5s - loss: 10.3298 - acc: 0.30 - ETA: 5s - loss: 10.3205 - acc: 0.30 - ETA: 5s - loss: 10.3302 - acc: 0.30 - ETA: 5s - loss: 10.3419 - acc: 0.30 - ETA: 5s - loss: 10.3094 - acc: 0.30 - ETA: 5s - loss: 10.3151 - acc: 0.30 - ETA: 5s - loss: 10.3182 - acc: 0.30 - ETA: 5s - loss: 10.3233 - acc: 0.30 - ETA: 5s - loss: 10.3392 - acc: 0.30 - ETA: 5s - loss: 10.3247 - acc: 0.30 - ETA: 5s - loss: 10.2825 - acc: 0.30 - ETA: 5s - loss: 10.2964 - acc: 0.30 - ETA: 5s - loss: 10.3169 - acc: 0.30 - ETA: 5s - loss: 10.2972 - acc: 0.30 - ETA: 5s - loss: 10.2956 - acc: 0.30 - ETA: 5s - loss: 10.2799 - acc: 0.30 - ETA: 5s - loss: 10.2937 - acc: 0.30 - ETA: 5s - loss: 10.2983 - acc: 0.30 - ETA: 5s - loss: 10.3128 - acc: 0.30 - ETA: 5s - loss: 10.3060 - acc: 0.30 - ETA: 5s - loss: 10.2838 - acc: 0.30 - ETA: 5s - loss: 10.2830 - acc: 0.30 - ETA: 5s - loss: 10.2797 - acc: 0.30 - ETA: 5s - loss: 10.2590 - acc: 0.30 - ETA: 5s - loss: 10.2781 - acc: 0.30 - ETA: 5s - loss: 10.2987 - acc: 0.30 - ETA: 4s - loss: 10.2681 - acc: 0.30 - ETA: 4s - loss: 10.2573 - acc: 0.30 - ETA: 4s - loss: 10.2555 - acc: 0.30 - ETA: 4s - loss: 10.2553 - acc: 0.30 - ETA: 4s - loss: 10.2537 - acc: 0.30 - ETA: 4s - loss: 10.2674 - acc: 0.30 - ETA: 4s - loss: 10.2497 - acc: 0.30 - ETA: 4s - loss: 10.2530 - acc: 0.30 - ETA: 4s - loss: 10.2544 - acc: 0.30 - ETA: 4s - loss: 10.2371 - acc: 0.30 - ETA: 4s - loss: 10.2207 - acc: 0.30 - ETA: 3s - loss: 10.2320 - acc: 0.30 - ETA: 3s - loss: 10.2314 - acc: 0.30 - ETA: 3s - loss: 10.2226 - acc: 0.30 - ETA: 3s - loss: 10.2269 - acc: 0.31 - ETA: 3s - loss: 10.2361 - acc: 0.30 - ETA: 3s - loss: 10.2419 - acc: 0.30 - ETA: 3s - loss: 10.2416 - acc: 0.30 - ETA: 3s - loss: 10.2437 - acc: 0.30 - ETA: 3s - loss: 10.2452 - acc: 0.30 - ETA: 3s - loss: 10.2427 - acc: 0.30 - ETA: 3s - loss: 10.2305 - acc: 0.30 - ETA: 3s - loss: 10.2427 - acc: 0.30 - ETA: 3s - loss: 10.2432 - acc: 0.30 - ETA: 3s - loss: 10.2199 - acc: 0.31 - ETA: 3s - loss: 10.2247 - acc: 0.30 - ETA: 3s - loss: 10.2217 - acc: 0.30 - ETA: 3s - loss: 10.2261 - acc: 0.31 - ETA: 3s - loss: 10.2094 - acc: 0.31 - ETA: 2s - loss: 10.2105 - acc: 0.31 - ETA: 2s - loss: 10.1979 - acc: 0.31 - ETA: 2s - loss: 10.1993 - acc: 0.31 - ETA: 2s - loss: 10.1998 - acc: 0.31 - ETA: 2s - loss: 10.2020 - acc: 0.31 - ETA: 2s - loss: 10.1861 - acc: 0.31 - ETA: 2s - loss: 10.1823 - acc: 0.31 - ETA: 2s - loss: 10.1695 - acc: 0.31 - ETA: 2s - loss: 10.1788 - acc: 0.31 - ETA: 2s - loss: 10.1802 - acc: 0.31 - ETA: 2s - loss: 10.2035 - acc: 0.31 - ETA: 2s - loss: 10.2061 - acc: 0.31 - ETA: 1s - loss: 10.1979 - acc: 0.31 - ETA: 1s - loss: 10.1973 - acc: 0.31 - ETA: 1s - loss: 10.1972 - acc: 0.31 - ETA: 1s - loss: 10.1964 - acc: 0.31 - ETA: 1s - loss: 10.2009 - acc: 0.31 - ETA: 1s - loss: 10.2031 - acc: 0.31 - ETA: 1s - loss: 10.1946 - acc: 0.31 - ETA: 1s - loss: 10.2066 - acc: 0.31 - ETA: 1s - loss: 10.2080 - acc: 0.31 - ETA: 1s - loss: 10.2126 - acc: 0.31 - ETA: 1s - loss: 10.2193 - acc: 0.31 - ETA: 1s - loss: 10.2301 - acc: 0.31 - ETA: 1s - loss: 10.2235 - acc: 0.31 - ETA: 1s - loss: 10.2109 - acc: 0.31 - ETA: 1s - loss: 10.2147 - acc: 0.31 - ETA: 1s - loss: 10.2225 - acc: 0.31 - ETA: 1s - loss: 10.2203 - acc: 0.31 - ETA: 1s - loss: 10.2273 - acc: 0.31 - ETA: 1s - loss: 10.2207 - acc: 0.31 - ETA: 0s - loss: 10.2120 - acc: 0.31 - ETA: 0s - loss: 10.2230 - acc: 0.31 - ETA: 0s - loss: 10.2160 - acc: 0.31 - ETA: 0s - loss: 10.2336 - acc: 0.31 - ETA: 0s - loss: 10.2381 - acc: 0.31 - ETA: 0s - loss: 10.2212 - acc: 0.31 - ETA: 0s - loss: 10.2245 - acc: 0.31 - ETA: 0s - loss: 10.2298 - acc: 0.31 - ETA: 0s - loss: 10.2164 - acc: 0.31 - ETA: 0s - loss: 10.2115 - acc: 0.31 - ETA: 0s - loss: 10.2001 - acc: 0.31 - ETA: 0s - loss: 10.1873 - acc: 0.3143Epoch 00002: val_loss improved from 10.56121 to 10.32637, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 10s - loss: 10.1985 - acc: 0.3135 - val_loss: 10.3264 - val_acc: 0.2982 Epoch 4/20 6640/6680 [============================>.] - ETA: 17s - loss: 10.6145 - acc: 0.300 - ETA: 8s - loss: 9.1817 - acc: 0.3750 - ETA: 7s - loss: 8.8921 - acc: 0.385 - ETA: 7s - loss: 9.4291 - acc: 0.355 - ETA: 6s - loss: 9.6487 - acc: 0.350 - ETA: 6s - loss: 9.7563 - acc: 0.346 - ETA: 6s - loss: 9.7938 - acc: 0.344 - ETA: 6s - loss: 9.5314 - acc: 0.365 - ETA: 6s - loss: 9.4424 - acc: 0.372 - ETA: 5s - loss: 9.3115 - acc: 0.378 - ETA: 5s - loss: 9.4538 - acc: 0.372 - ETA: 5s - loss: 9.6822 - acc: 0.358 - ETA: 5s - loss: 9.7007 - acc: 0.356 - ETA: 5s - loss: 9.7663 - acc: 0.348 - ETA: 5s - loss: 9.7648 - acc: 0.348 - ETA: 5s - loss: 9.7439 - acc: 0.352 - ETA: 5s - loss: 9.7605 - acc: 0.351 - ETA: 5s - loss: 9.7701 - acc: 0.351 - ETA: 5s - loss: 9.6623 - acc: 0.355 - ETA: 5s - loss: 9.6969 - acc: 0.352 - ETA: 5s - loss: 9.6562 - acc: 0.356 - ETA: 5s - loss: 9.6481 - acc: 0.358 - ETA: 5s - loss: 9.7280 - acc: 0.353 - ETA: 5s - loss: 9.7195 - acc: 0.353 - ETA: 5s - loss: 9.7265 - acc: 0.353 - ETA: 5s - loss: 9.7496 - acc: 0.352 - ETA: 5s - loss: 9.7803 - acc: 0.351 - ETA: 5s - loss: 9.7973 - acc: 0.350 - ETA: 5s - loss: 9.7913 - acc: 0.350 - ETA: 6s - loss: 9.7862 - acc: 0.350 - ETA: 6s - loss: 9.7642 - acc: 0.351 - ETA: 6s - loss: 9.7908 - acc: 0.349 - ETA: 6s - loss: 9.8097 - acc: 0.348 - ETA: 6s - loss: 9.7953 - acc: 0.349 - ETA: 6s - loss: 9.7959 - acc: 0.349 - ETA: 6s - loss: 9.7901 - acc: 0.349 - ETA: 6s - loss: 9.8011 - acc: 0.349 - ETA: 6s - loss: 9.7904 - acc: 0.349 - ETA: 6s - loss: 9.7629 - acc: 0.351 - ETA: 6s - loss: 9.8381 - acc: 0.347 - ETA: 6s - loss: 9.8856 - acc: 0.344 - ETA: 6s - loss: 9.8652 - acc: 0.346 - ETA: 6s - loss: 9.8400 - acc: 0.348 - ETA: 5s - loss: 9.8887 - acc: 0.344 - ETA: 5s - loss: 9.8877 - acc: 0.344 - ETA: 5s - loss: 9.8859 - acc: 0.344 - ETA: 5s - loss: 9.9271 - acc: 0.340 - ETA: 5s - loss: 9.9169 - acc: 0.341 - ETA: 5s - loss: 9.8885 - acc: 0.341 - ETA: 5s - loss: 9.8705 - acc: 0.341 - ETA: 5s - loss: 9.8848 - acc: 0.341 - ETA: 5s - loss: 9.8779 - acc: 0.342 - ETA: 5s - loss: 9.8704 - acc: 0.342 - ETA: 5s - loss: 9.8714 - acc: 0.342 - ETA: 5s - loss: 9.8401 - acc: 0.344 - ETA: 5s - loss: 9.8467 - acc: 0.343 - ETA: 5s - loss: 9.8575 - acc: 0.343 - ETA: 5s - loss: 9.8565 - acc: 0.343 - ETA: 5s - loss: 9.8587 - acc: 0.343 - ETA: 5s - loss: 9.8515 - acc: 0.343 - ETA: 5s - loss: 9.8560 - acc: 0.343 - ETA: 5s - loss: 9.8471 - acc: 0.344 - ETA: 5s - loss: 9.8494 - acc: 0.344 - ETA: 5s - loss: 9.8513 - acc: 0.342 - ETA: 5s - loss: 9.8515 - acc: 0.343 - ETA: 5s - loss: 9.8469 - acc: 0.344 - ETA: 5s - loss: 9.8100 - acc: 0.346 - ETA: 5s - loss: 9.7921 - acc: 0.346 - ETA: 5s - loss: 9.8132 - acc: 0.345 - ETA: 5s - loss: 9.8192 - acc: 0.344 - ETA: 4s - loss: 9.8185 - acc: 0.344 - ETA: 4s - loss: 9.8491 - acc: 0.342 - ETA: 5s - loss: 9.8439 - acc: 0.342 - ETA: 5s - loss: 9.8251 - acc: 0.343 - ETA: 5s - loss: 9.8383 - acc: 0.343 - ETA: 5s - loss: 9.8186 - acc: 0.343 - ETA: 5s - loss: 9.8125 - acc: 0.343 - ETA: 5s - loss: 9.8165 - acc: 0.343 - ETA: 5s - loss: 9.8190 - acc: 0.343 - ETA: 5s - loss: 9.8356 - acc: 0.342 - ETA: 5s - loss: 9.8467 - acc: 0.340 - ETA: 5s - loss: 9.8547 - acc: 0.340 - ETA: 4s - loss: 9.8585 - acc: 0.339 - ETA: 4s - loss: 9.8675 - acc: 0.339 - ETA: 4s - loss: 9.8674 - acc: 0.339 - ETA: 4s - loss: 9.8796 - acc: 0.338 - ETA: 4s - loss: 9.8673 - acc: 0.338 - ETA: 4s - loss: 9.8662 - acc: 0.338 - ETA: 4s - loss: 9.8479 - acc: 0.339 - ETA: 4s - loss: 9.8470 - acc: 0.340 - ETA: 4s - loss: 9.8421 - acc: 0.340 - ETA: 4s - loss: 9.8456 - acc: 0.340 - ETA: 4s - loss: 9.8330 - acc: 0.341 - ETA: 4s - loss: 9.8054 - acc: 0.343 - ETA: 4s - loss: 9.8165 - acc: 0.343 - ETA: 4s - loss: 9.7995 - acc: 0.343 - ETA: 4s - loss: 9.7894 - acc: 0.343 - ETA: 4s - loss: 9.7758 - acc: 0.343 - ETA: 4s - loss: 9.8035 - acc: 0.342 - ETA: 3s - loss: 9.7840 - acc: 0.342 - ETA: 3s - loss: 9.7810 - acc: 0.343 - ETA: 3s - loss: 9.7926 - acc: 0.342 - ETA: 3s - loss: 9.8015 - acc: 0.342 - ETA: 3s - loss: 9.7955 - acc: 0.342 - ETA: 3s - loss: 9.7980 - acc: 0.343 - ETA: 3s - loss: 9.7976 - acc: 0.343 - ETA: 3s - loss: 9.8180 - acc: 0.341 - ETA: 3s - loss: 9.8295 - acc: 0.341 - ETA: 3s - loss: 9.8334 - acc: 0.340 - ETA: 2s - loss: 9.8364 - acc: 0.339 - ETA: 2s - loss: 9.8353 - acc: 0.340 - ETA: 2s - loss: 9.8289 - acc: 0.340 - ETA: 2s - loss: 9.8235 - acc: 0.340 - ETA: 2s - loss: 9.8228 - acc: 0.340 - ETA: 2s - loss: 9.8084 - acc: 0.340 - ETA: 2s - loss: 9.8016 - acc: 0.341 - ETA: 2s - loss: 9.8044 - acc: 0.341 - ETA: 2s - loss: 9.8036 - acc: 0.341 - ETA: 2s - loss: 9.8069 - acc: 0.341 - ETA: 2s - loss: 9.8190 - acc: 0.341 - ETA: 2s - loss: 9.8131 - acc: 0.341 - ETA: 2s - loss: 9.7958 - acc: 0.342 - ETA: 2s - loss: 9.7995 - acc: 0.341 - ETA: 2s - loss: 9.7925 - acc: 0.342 - ETA: 2s - loss: 9.8063 - acc: 0.341 - ETA: 2s - loss: 9.7954 - acc: 0.342 - ETA: 1s - loss: 9.7940 - acc: 0.342 - ETA: 1s - loss: 9.7842 - acc: 0.343 - ETA: 1s - loss: 9.7934 - acc: 0.343 - ETA: 1s - loss: 9.7921 - acc: 0.344 - ETA: 1s - loss: 9.7811 - acc: 0.344 - ETA: 1s - loss: 9.7936 - acc: 0.344 - ETA: 1s - loss: 9.8056 - acc: 0.343 - ETA: 1s - loss: 9.8148 - acc: 0.342 - ETA: 1s - loss: 9.8272 - acc: 0.341 - ETA: 1s - loss: 9.8094 - acc: 0.343 - ETA: 1s - loss: 9.8300 - acc: 0.342 - ETA: 0s - loss: 9.8239 - acc: 0.342 - ETA: 0s - loss: 9.8312 - acc: 0.342 - ETA: 0s - loss: 9.8276 - acc: 0.342 - ETA: 0s - loss: 9.8319 - acc: 0.342 - ETA: 0s - loss: 9.8376 - acc: 0.342 - ETA: 0s - loss: 9.8341 - acc: 0.342 - ETA: 0s - loss: 9.8377 - acc: 0.342 - ETA: 0s - loss: 9.8329 - acc: 0.343 - ETA: 0s - loss: 9.8324 - acc: 0.343 - ETA: 0s - loss: 9.8420 - acc: 0.342 - ETA: 0s - loss: 9.8418 - acc: 0.343 - ETA: 0s - loss: 9.8413 - acc: 0.343 - ETA: 0s - loss: 9.8349 - acc: 0.343 - ETA: 0s - loss: 9.8319 - acc: 0.343 - ETA: 0s - loss: 9.8335 - acc: 0.343 - ETA: 0s - loss: 9.8314 - acc: 0.343 - ETA: 0s - loss: 9.8261 - acc: 0.344 - ETA: 0s - loss: 9.8282 - acc: 0.344 - ETA: 0s - loss: 9.8284 - acc: 0.344 - ETA: 0s - loss: 9.8291 - acc: 0.343 - ETA: 0s - loss: 9.8389 - acc: 0.3435Epoch 00003: val_loss improved from 10.32637 to 10.01467, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 11s - loss: 9.8458 - acc: 0.3433 - val_loss: 10.0147 - val_acc: 0.3150 Epoch 5/20 6660/6680 [============================>.] - ETA: 31s - loss: 8.8116 - acc: 0.40 - ETA: 13s - loss: 10.0407 - acc: 0.350 - ETA: 10s - loss: 9.5782 - acc: 0.385 - ETA: 8s - loss: 9.6212 - acc: 0.3850 - ETA: 8s - loss: 9.8871 - acc: 0.369 - ETA: 7s - loss: 9.5088 - acc: 0.390 - ETA: 7s - loss: 9.6253 - acc: 0.384 - ETA: 7s - loss: 9.7163 - acc: 0.377 - ETA: 6s - loss: 9.5703 - acc: 0.386 - ETA: 6s - loss: 9.6810 - acc: 0.378 - ETA: 6s - loss: 9.7469 - acc: 0.375 - ETA: 6s - loss: 9.7491 - acc: 0.375 - ETA: 6s - loss: 9.8603 - acc: 0.368 - ETA: 7s - loss: 9.8679 - acc: 0.367 - ETA: 7s - loss: 9.7634 - acc: 0.371 - ETA: 7s - loss: 9.7455 - acc: 0.372 - ETA: 7s - loss: 9.6220 - acc: 0.380 - ETA: 8s - loss: 9.6263 - acc: 0.379 - ETA: 8s - loss: 9.5979 - acc: 0.379 - ETA: 9s - loss: 9.5742 - acc: 0.380 - ETA: 9s - loss: 9.4958 - acc: 0.384 - ETA: 8s - loss: 9.4678 - acc: 0.386 - ETA: 8s - loss: 9.5082 - acc: 0.385 - ETA: 8s - loss: 9.5475 - acc: 0.384 - ETA: 8s - loss: 9.4006 - acc: 0.392 - ETA: 7s - loss: 9.3462 - acc: 0.395 - ETA: 7s - loss: 9.3625 - acc: 0.392 - ETA: 7s - loss: 9.3563 - acc: 0.393 - ETA: 7s - loss: 9.3723 - acc: 0.392 - ETA: 7s - loss: 9.3481 - acc: 0.394 - ETA: 7s - loss: 9.3562 - acc: 0.392 - ETA: 6s - loss: 9.3787 - acc: 0.390 - ETA: 6s - loss: 9.4223 - acc: 0.386 - ETA: 6s - loss: 9.4503 - acc: 0.382 - ETA: 6s - loss: 9.4604 - acc: 0.382 - ETA: 6s - loss: 9.4921 - acc: 0.380 - ETA: 6s - loss: 9.4990 - acc: 0.380 - ETA: 6s - loss: 9.5216 - acc: 0.379 - ETA: 6s - loss: 9.5322 - acc: 0.377 - ETA: 6s - loss: 9.5479 - acc: 0.377 - ETA: 6s - loss: 9.5289 - acc: 0.378 - ETA: 5s - loss: 9.5342 - acc: 0.377 - ETA: 5s - loss: 9.5405 - acc: 0.377 - ETA: 5s - loss: 9.5278 - acc: 0.378 - ETA: 5s - loss: 9.5244 - acc: 0.378 - ETA: 5s - loss: 9.5145 - acc: 0.378 - ETA: 5s - loss: 9.5078 - acc: 0.378 - ETA: 5s - loss: 9.5300 - acc: 0.377 - ETA: 5s - loss: 9.5519 - acc: 0.376 - ETA: 6s - loss: 9.5396 - acc: 0.377 - ETA: 6s - loss: 9.5133 - acc: 0.379 - ETA: 6s - loss: 9.4745 - acc: 0.381 - ETA: 6s - loss: 9.4658 - acc: 0.382 - ETA: 6s - loss: 9.4675 - acc: 0.382 - ETA: 6s - loss: 9.4634 - acc: 0.382 - ETA: 6s - loss: 9.4544 - acc: 0.383 - ETA: 6s - loss: 9.4544 - acc: 0.383 - ETA: 6s - loss: 9.4547 - acc: 0.382 - ETA: 6s - loss: 9.4631 - acc: 0.382 - ETA: 6s - loss: 9.4616 - acc: 0.382 - ETA: 6s - loss: 9.4329 - acc: 0.383 - ETA: 6s - loss: 9.4481 - acc: 0.383 - ETA: 6s - loss: 9.4297 - acc: 0.384 - ETA: 5s - loss: 9.4227 - acc: 0.385 - ETA: 5s - loss: 9.4210 - acc: 0.385 - ETA: 5s - loss: 9.4346 - acc: 0.384 - ETA: 5s - loss: 9.4377 - acc: 0.384 - ETA: 5s - loss: 9.4207 - acc: 0.385 - ETA: 5s - loss: 9.4297 - acc: 0.384 - ETA: 5s - loss: 9.4386 - acc: 0.384 - ETA: 5s - loss: 9.4569 - acc: 0.383 - ETA: 5s - loss: 9.4858 - acc: 0.381 - ETA: 5s - loss: 9.4869 - acc: 0.381 - ETA: 5s - loss: 9.5125 - acc: 0.380 - ETA: 4s - loss: 9.5204 - acc: 0.380 - ETA: 4s - loss: 9.5083 - acc: 0.381 - ETA: 4s - loss: 9.5111 - acc: 0.381 - ETA: 4s - loss: 9.5041 - acc: 0.382 - ETA: 4s - loss: 9.5152 - acc: 0.381 - ETA: 4s - loss: 9.5445 - acc: 0.379 - ETA: 4s - loss: 9.5169 - acc: 0.381 - ETA: 4s - loss: 9.5231 - acc: 0.381 - ETA: 4s - loss: 9.5180 - acc: 0.381 - ETA: 4s - loss: 9.4912 - acc: 0.382 - ETA: 4s - loss: 9.4974 - acc: 0.382 - ETA: 4s - loss: 9.5212 - acc: 0.381 - ETA: 4s - loss: 9.5241 - acc: 0.381 - ETA: 4s - loss: 9.5207 - acc: 0.381 - ETA: 4s - loss: 9.5130 - acc: 0.381 - ETA: 4s - loss: 9.5097 - acc: 0.382 - ETA: 4s - loss: 9.5201 - acc: 0.381 - ETA: 4s - loss: 9.5210 - acc: 0.381 - ETA: 4s - loss: 9.4987 - acc: 0.382 - ETA: 4s - loss: 9.5039 - acc: 0.381 - ETA: 4s - loss: 9.5141 - acc: 0.381 - ETA: 4s - loss: 9.5234 - acc: 0.380 - ETA: 4s - loss: 9.5222 - acc: 0.380 - ETA: 4s - loss: 9.5309 - acc: 0.379 - ETA: 4s - loss: 9.5246 - acc: 0.380 - ETA: 4s - loss: 9.5330 - acc: 0.379 - ETA: 4s - loss: 9.5380 - acc: 0.379 - ETA: 4s - loss: 9.5496 - acc: 0.378 - ETA: 3s - loss: 9.5477 - acc: 0.379 - ETA: 3s - loss: 9.5494 - acc: 0.379 - ETA: 3s - loss: 9.5554 - acc: 0.379 - ETA: 3s - loss: 9.5704 - acc: 0.378 - ETA: 3s - loss: 9.5853 - acc: 0.377 - ETA: 3s - loss: 9.5940 - acc: 0.377 - ETA: 3s - loss: 9.5877 - acc: 0.377 - ETA: 3s - loss: 9.6007 - acc: 0.377 - ETA: 3s - loss: 9.6221 - acc: 0.376 - ETA: 3s - loss: 9.6356 - acc: 0.375 - ETA: 3s - loss: 9.6160 - acc: 0.376 - ETA: 2s - loss: 9.6168 - acc: 0.376 - ETA: 2s - loss: 9.6040 - acc: 0.377 - ETA: 2s - loss: 9.6210 - acc: 0.376 - ETA: 2s - loss: 9.6257 - acc: 0.376 - ETA: 2s - loss: 9.6200 - acc: 0.377 - ETA: 2s - loss: 9.6274 - acc: 0.376 - ETA: 2s - loss: 9.6187 - acc: 0.376 - ETA: 2s - loss: 9.6160 - acc: 0.377 - ETA: 2s - loss: 9.6335 - acc: 0.375 - ETA: 2s - loss: 9.6500 - acc: 0.374 - ETA: 2s - loss: 9.6419 - acc: 0.375 - ETA: 2s - loss: 9.6485 - acc: 0.375 - ETA: 2s - loss: 9.6433 - acc: 0.375 - ETA: 2s - loss: 9.6466 - acc: 0.375 - ETA: 2s - loss: 9.6570 - acc: 0.375 - ETA: 1s - loss: 9.6797 - acc: 0.373 - ETA: 1s - loss: 9.6563 - acc: 0.374 - ETA: 1s - loss: 9.6460 - acc: 0.375 - ETA: 1s - loss: 9.6405 - acc: 0.376 - ETA: 1s - loss: 9.6576 - acc: 0.374 - ETA: 1s - loss: 9.6609 - acc: 0.374 - ETA: 1s - loss: 9.6715 - acc: 0.373 - ETA: 1s - loss: 9.6772 - acc: 0.372 - ETA: 1s - loss: 9.6730 - acc: 0.373 - ETA: 1s - loss: 9.6729 - acc: 0.372 - ETA: 1s - loss: 9.6701 - acc: 0.372 - ETA: 1s - loss: 9.6676 - acc: 0.373 - ETA: 1s - loss: 9.6632 - acc: 0.373 - ETA: 0s - loss: 9.6572 - acc: 0.373 - ETA: 0s - loss: 9.6616 - acc: 0.373 - ETA: 0s - loss: 9.6676 - acc: 0.372 - ETA: 0s - loss: 9.6605 - acc: 0.373 - ETA: 0s - loss: 9.6644 - acc: 0.373 - ETA: 0s - loss: 9.6773 - acc: 0.372 - ETA: 0s - loss: 9.6825 - acc: 0.372 - ETA: 0s - loss: 9.6932 - acc: 0.371 - ETA: 0s - loss: 9.6885 - acc: 0.371 - ETA: 0s - loss: 9.7004 - acc: 0.370 - ETA: 0s - loss: 9.7014 - acc: 0.370 - ETA: 0s - loss: 9.6963 - acc: 0.371 - ETA: 0s - loss: 9.6889 - acc: 0.371 - ETA: 0s - loss: 9.6822 - acc: 0.372 - ETA: 0s - loss: 9.6822 - acc: 0.372 - ETA: 0s - loss: 9.6798 - acc: 0.372 - ETA: 0s - loss: 9.6725 - acc: 0.372 - ETA: 0s - loss: 9.6774 - acc: 0.372 - ETA: 0s - loss: 9.6799 - acc: 0.372 - ETA: 0s - loss: 9.6899 - acc: 0.3718Epoch 00004: val_loss improved from 10.01467 to 9.93790, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 9.6923 - acc: 0.3717 - val_loss: 9.9379 - val_acc: 0.3389 Epoch 6/20 6660/6680 [============================>.] - ETA: 8s - loss: 8.0601 - acc: 0.500 - ETA: 5s - loss: 10.9646 - acc: 0.31 - ETA: 6s - loss: 10.2046 - acc: 0.35 - ETA: 6s - loss: 9.9302 - acc: 0.3727 - ETA: 6s - loss: 9.5526 - acc: 0.396 - ETA: 6s - loss: 9.6875 - acc: 0.388 - ETA: 6s - loss: 9.6623 - acc: 0.390 - ETA: 5s - loss: 9.7423 - acc: 0.384 - ETA: 5s - loss: 9.6428 - acc: 0.392 - ETA: 5s - loss: 9.6480 - acc: 0.386 - ETA: 5s - loss: 9.7535 - acc: 0.381 - ETA: 5s - loss: 9.6777 - acc: 0.386 - ETA: 5s - loss: 9.6623 - acc: 0.387 - ETA: 5s - loss: 9.6047 - acc: 0.391 - ETA: 5s - loss: 9.7036 - acc: 0.386 - ETA: 6s - loss: 9.6552 - acc: 0.388 - ETA: 6s - loss: 9.7255 - acc: 0.385 - ETA: 6s - loss: 9.6914 - acc: 0.387 - ETA: 6s - loss: 9.6745 - acc: 0.388 - ETA: 7s - loss: 9.7229 - acc: 0.386 - ETA: 7s - loss: 9.7237 - acc: 0.385 - ETA: 7s - loss: 9.7418 - acc: 0.382 - ETA: 8s - loss: 9.7405 - acc: 0.383 - ETA: 8s - loss: 9.6945 - acc: 0.386 - ETA: 8s - loss: 9.6951 - acc: 0.386 - ETA: 8s - loss: 9.6756 - acc: 0.386 - ETA: 8s - loss: 9.6756 - acc: 0.386 - ETA: 8s - loss: 9.7036 - acc: 0.385 - ETA: 9s - loss: 9.7032 - acc: 0.385 - ETA: 9s - loss: 9.7163 - acc: 0.385 - ETA: 9s - loss: 9.7110 - acc: 0.384 - ETA: 8s - loss: 9.6443 - acc: 0.387 - ETA: 8s - loss: 9.6950 - acc: 0.383 - ETA: 8s - loss: 9.7054 - acc: 0.383 - ETA: 8s - loss: 9.7484 - acc: 0.379 - ETA: 8s - loss: 9.8140 - acc: 0.374 - ETA: 8s - loss: 9.7806 - acc: 0.376 - ETA: 7s - loss: 9.7551 - acc: 0.378 - ETA: 7s - loss: 9.7634 - acc: 0.378 - ETA: 7s - loss: 9.8266 - acc: 0.375 - ETA: 7s - loss: 9.8834 - acc: 0.369 - ETA: 7s - loss: 9.8419 - acc: 0.371 - ETA: 7s - loss: 9.8406 - acc: 0.372 - ETA: 7s - loss: 9.7457 - acc: 0.378 - ETA: 6s - loss: 9.7673 - acc: 0.375 - ETA: 6s - loss: 9.7696 - acc: 0.375 - ETA: 6s - loss: 9.7992 - acc: 0.373 - ETA: 6s - loss: 9.7647 - acc: 0.375 - ETA: 6s - loss: 9.7296 - acc: 0.376 - ETA: 6s - loss: 9.7000 - acc: 0.378 - ETA: 6s - loss: 9.6625 - acc: 0.379 - ETA: 6s - loss: 9.6576 - acc: 0.379 - ETA: 6s - loss: 9.6526 - acc: 0.380 - ETA: 6s - loss: 9.6752 - acc: 0.378 - ETA: 5s - loss: 9.6546 - acc: 0.379 - ETA: 5s - loss: 9.6372 - acc: 0.380 - ETA: 5s - loss: 9.6382 - acc: 0.380 - ETA: 5s - loss: 9.6449 - acc: 0.379 - ETA: 5s - loss: 9.6417 - acc: 0.379 - ETA: 5s - loss: 9.6202 - acc: 0.380 - ETA: 5s - loss: 9.5909 - acc: 0.382 - ETA: 5s - loss: 9.5889 - acc: 0.381 - ETA: 5s - loss: 9.6061 - acc: 0.380 - ETA: 5s - loss: 9.5685 - acc: 0.382 - ETA: 5s - loss: 9.5468 - acc: 0.384 - ETA: 5s - loss: 9.5757 - acc: 0.382 - ETA: 5s - loss: 9.5791 - acc: 0.381 - ETA: 5s - loss: 9.5800 - acc: 0.381 - ETA: 5s - loss: 9.5645 - acc: 0.382 - ETA: 5s - loss: 9.5654 - acc: 0.382 - ETA: 5s - loss: 9.5618 - acc: 0.382 - ETA: 5s - loss: 9.5481 - acc: 0.383 - ETA: 5s - loss: 9.5394 - acc: 0.383 - ETA: 5s - loss: 9.5250 - acc: 0.383 - ETA: 5s - loss: 9.5379 - acc: 0.383 - ETA: 5s - loss: 9.5421 - acc: 0.382 - ETA: 5s - loss: 9.5335 - acc: 0.383 - ETA: 5s - loss: 9.5614 - acc: 0.381 - ETA: 5s - loss: 9.5589 - acc: 0.381 - ETA: 5s - loss: 9.5665 - acc: 0.381 - ETA: 5s - loss: 9.5627 - acc: 0.381 - ETA: 5s - loss: 9.5542 - acc: 0.381 - ETA: 5s - loss: 9.5526 - acc: 0.381 - ETA: 5s - loss: 9.5636 - acc: 0.380 - ETA: 5s - loss: 9.5549 - acc: 0.381 - ETA: 4s - loss: 9.5607 - acc: 0.381 - ETA: 4s - loss: 9.5319 - acc: 0.382 - ETA: 4s - loss: 9.5266 - acc: 0.383 - ETA: 4s - loss: 9.5327 - acc: 0.383 - ETA: 4s - loss: 9.5533 - acc: 0.381 - ETA: 4s - loss: 9.5611 - acc: 0.381 - ETA: 4s - loss: 9.5381 - acc: 0.382 - ETA: 4s - loss: 9.5441 - acc: 0.382 - ETA: 4s - loss: 9.5501 - acc: 0.381 - ETA: 4s - loss: 9.5392 - acc: 0.382 - ETA: 3s - loss: 9.5255 - acc: 0.382 - ETA: 3s - loss: 9.5103 - acc: 0.383 - ETA: 3s - loss: 9.5148 - acc: 0.383 - ETA: 3s - loss: 9.5196 - acc: 0.383 - ETA: 3s - loss: 9.4970 - acc: 0.384 - ETA: 3s - loss: 9.4900 - acc: 0.384 - ETA: 3s - loss: 9.5027 - acc: 0.383 - ETA: 3s - loss: 9.5021 - acc: 0.383 - ETA: 3s - loss: 9.5221 - acc: 0.382 - ETA: 3s - loss: 9.5351 - acc: 0.381 - ETA: 3s - loss: 9.5424 - acc: 0.380 - ETA: 3s - loss: 9.5443 - acc: 0.380 - ETA: 3s - loss: 9.5448 - acc: 0.380 - ETA: 3s - loss: 9.5545 - acc: 0.380 - ETA: 3s - loss: 9.5560 - acc: 0.379 - ETA: 3s - loss: 9.5566 - acc: 0.379 - ETA: 3s - loss: 9.5363 - acc: 0.381 - ETA: 3s - loss: 9.5406 - acc: 0.380 - ETA: 3s - loss: 9.5412 - acc: 0.381 - ETA: 3s - loss: 9.5461 - acc: 0.380 - ETA: 3s - loss: 9.5387 - acc: 0.380 - ETA: 3s - loss: 9.5303 - acc: 0.381 - ETA: 3s - loss: 9.5208 - acc: 0.381 - ETA: 3s - loss: 9.5214 - acc: 0.381 - ETA: 3s - loss: 9.5305 - acc: 0.380 - ETA: 3s - loss: 9.5362 - acc: 0.380 - ETA: 2s - loss: 9.5357 - acc: 0.380 - ETA: 2s - loss: 9.5292 - acc: 0.381 - ETA: 2s - loss: 9.5112 - acc: 0.382 - ETA: 2s - loss: 9.5129 - acc: 0.381 - ETA: 2s - loss: 9.5000 - acc: 0.382 - ETA: 2s - loss: 9.4895 - acc: 0.383 - ETA: 2s - loss: 9.4883 - acc: 0.383 - ETA: 2s - loss: 9.4913 - acc: 0.382 - ETA: 2s - loss: 9.4991 - acc: 0.381 - ETA: 2s - loss: 9.5013 - acc: 0.381 - ETA: 2s - loss: 9.5207 - acc: 0.380 - ETA: 2s - loss: 9.5169 - acc: 0.380 - ETA: 2s - loss: 9.5064 - acc: 0.381 - ETA: 1s - loss: 9.5096 - acc: 0.381 - ETA: 1s - loss: 9.5027 - acc: 0.381 - ETA: 1s - loss: 9.5071 - acc: 0.381 - ETA: 1s - loss: 9.5142 - acc: 0.380 - ETA: 1s - loss: 9.5146 - acc: 0.380 - ETA: 1s - loss: 9.5253 - acc: 0.379 - ETA: 1s - loss: 9.5250 - acc: 0.379 - ETA: 1s - loss: 9.5198 - acc: 0.379 - ETA: 1s - loss: 9.5185 - acc: 0.379 - ETA: 1s - loss: 9.5199 - acc: 0.379 - ETA: 1s - loss: 9.5305 - acc: 0.378 - ETA: 1s - loss: 9.5234 - acc: 0.379 - ETA: 1s - loss: 9.5187 - acc: 0.379 - ETA: 0s - loss: 9.5144 - acc: 0.379 - ETA: 0s - loss: 9.5166 - acc: 0.379 - ETA: 0s - loss: 9.5257 - acc: 0.379 - ETA: 0s - loss: 9.5146 - acc: 0.379 - ETA: 0s - loss: 9.5208 - acc: 0.379 - ETA: 0s - loss: 9.5349 - acc: 0.378 - ETA: 0s - loss: 9.5309 - acc: 0.378 - ETA: 0s - loss: 9.5327 - acc: 0.378 - ETA: 0s - loss: 9.5306 - acc: 0.378 - ETA: 0s - loss: 9.5336 - acc: 0.378 - ETA: 0s - loss: 9.5223 - acc: 0.379 - ETA: 0s - loss: 9.5202 - acc: 0.379 - ETA: 0s - loss: 9.5232 - acc: 0.379 - ETA: 0s - loss: 9.5205 - acc: 0.379 - ETA: 0s - loss: 9.5104 - acc: 0.380 - ETA: 0s - loss: 9.5060 - acc: 0.380 - ETA: 0s - loss: 9.5046 - acc: 0.380 - ETA: 0s - loss: 9.4976 - acc: 0.380 - ETA: 0s - loss: 9.5070 - acc: 0.380 - ETA: 0s - loss: 9.5134 - acc: 0.379 - ETA: 0s - loss: 9.5156 - acc: 0.3794Epoch 00005: val_loss improved from 9.93790 to 9.85014, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 9.5185 - acc: 0.3793 - val_loss: 9.8501 - val_acc: 0.3222 Epoch 7/20 6640/6680 [============================>.] - ETA: 8s - loss: 8.4188 - acc: 0.450 - ETA: 6s - loss: 8.6297 - acc: 0.425 - ETA: 5s - loss: 8.8637 - acc: 0.414 - ETA: 5s - loss: 8.3907 - acc: 0.450 - ETA: 5s - loss: 8.0665 - acc: 0.476 - ETA: 5s - loss: 8.1927 - acc: 0.462 - ETA: 5s - loss: 8.2989 - acc: 0.460 - ETA: 5s - loss: 8.5079 - acc: 0.445 - ETA: 5s - loss: 8.5925 - acc: 0.442 - ETA: 5s - loss: 8.7882 - acc: 0.430 - ETA: 5s - loss: 8.7665 - acc: 0.433 - ETA: 5s - loss: 8.8030 - acc: 0.429 - ETA: 6s - loss: 8.8383 - acc: 0.427 - ETA: 6s - loss: 8.9100 - acc: 0.422 - ETA: 6s - loss: 8.9297 - acc: 0.423 - ETA: 6s - loss: 8.9925 - acc: 0.419 - ETA: 6s - loss: 9.0256 - acc: 0.418 - ETA: 6s - loss: 9.0993 - acc: 0.414 - ETA: 7s - loss: 9.0808 - acc: 0.415 - ETA: 7s - loss: 9.1674 - acc: 0.410 - ETA: 8s - loss: 9.1820 - acc: 0.408 - ETA: 8s - loss: 9.1929 - acc: 0.408 - ETA: 8s - loss: 9.2033 - acc: 0.408 - ETA: 9s - loss: 9.1631 - acc: 0.411 - ETA: 9s - loss: 9.1436 - acc: 0.412 - ETA: 9s - loss: 9.1380 - acc: 0.413 - ETA: 9s - loss: 9.1013 - acc: 0.415 - ETA: 9s - loss: 9.0735 - acc: 0.416 - ETA: 10s - loss: 9.0695 - acc: 0.41 - ETA: 9s - loss: 9.0246 - acc: 0.4170 - ETA: 9s - loss: 9.0219 - acc: 0.416 - ETA: 9s - loss: 9.0173 - acc: 0.417 - ETA: 9s - loss: 9.0677 - acc: 0.414 - ETA: 8s - loss: 9.0835 - acc: 0.412 - ETA: 8s - loss: 9.0597 - acc: 0.413 - ETA: 8s - loss: 9.1015 - acc: 0.410 - ETA: 8s - loss: 9.0174 - acc: 0.416 - ETA: 8s - loss: 8.9912 - acc: 0.418 - ETA: 8s - loss: 8.9919 - acc: 0.417 - ETA: 8s - loss: 8.9991 - acc: 0.418 - ETA: 7s - loss: 9.0335 - acc: 0.416 - ETA: 7s - loss: 9.0308 - acc: 0.417 - ETA: 7s - loss: 9.0445 - acc: 0.416 - ETA: 7s - loss: 9.0662 - acc: 0.415 - ETA: 7s - loss: 9.0994 - acc: 0.413 - ETA: 7s - loss: 9.0699 - acc: 0.415 - ETA: 7s - loss: 9.0589 - acc: 0.416 - ETA: 7s - loss: 9.1041 - acc: 0.412 - ETA: 6s - loss: 9.0795 - acc: 0.413 - ETA: 6s - loss: 9.1055 - acc: 0.411 - ETA: 6s - loss: 9.0936 - acc: 0.413 - ETA: 6s - loss: 9.1073 - acc: 0.411 - ETA: 6s - loss: 9.1398 - acc: 0.409 - ETA: 6s - loss: 9.1455 - acc: 0.409 - ETA: 6s - loss: 9.1271 - acc: 0.410 - ETA: 6s - loss: 9.1227 - acc: 0.411 - ETA: 6s - loss: 9.1252 - acc: 0.411 - ETA: 6s - loss: 9.1023 - acc: 0.412 - ETA: 6s - loss: 9.1118 - acc: 0.412 - ETA: 6s - loss: 9.1278 - acc: 0.411 - ETA: 6s - loss: 9.1387 - acc: 0.410 - ETA: 6s - loss: 9.1313 - acc: 0.410 - ETA: 6s - loss: 9.1041 - acc: 0.411 - ETA: 6s - loss: 9.1028 - acc: 0.411 - ETA: 6s - loss: 9.0895 - acc: 0.412 - ETA: 6s - loss: 9.0882 - acc: 0.412 - ETA: 6s - loss: 9.1004 - acc: 0.411 - ETA: 6s - loss: 9.0869 - acc: 0.412 - ETA: 6s - loss: 9.0735 - acc: 0.413 - ETA: 6s - loss: 9.0727 - acc: 0.413 - ETA: 6s - loss: 9.0712 - acc: 0.413 - ETA: 6s - loss: 9.0545 - acc: 0.414 - ETA: 6s - loss: 9.0600 - acc: 0.413 - ETA: 6s - loss: 9.0671 - acc: 0.413 - ETA: 6s - loss: 9.0490 - acc: 0.414 - ETA: 6s - loss: 9.0482 - acc: 0.414 - ETA: 6s - loss: 9.0581 - acc: 0.413 - ETA: 6s - loss: 9.0303 - acc: 0.414 - ETA: 6s - loss: 9.0209 - acc: 0.415 - ETA: 6s - loss: 9.0372 - acc: 0.414 - ETA: 6s - loss: 9.0403 - acc: 0.414 - ETA: 6s - loss: 9.0548 - acc: 0.413 - ETA: 5s - loss: 9.0629 - acc: 0.413 - ETA: 5s - loss: 9.0614 - acc: 0.412 - ETA: 5s - loss: 9.0445 - acc: 0.414 - ETA: 5s - loss: 9.0331 - acc: 0.414 - ETA: 5s - loss: 9.0420 - acc: 0.413 - ETA: 5s - loss: 9.0687 - acc: 0.412 - ETA: 5s - loss: 9.0866 - acc: 0.411 - ETA: 5s - loss: 9.1059 - acc: 0.410 - ETA: 5s - loss: 9.1292 - acc: 0.409 - ETA: 5s - loss: 9.1251 - acc: 0.409 - ETA: 5s - loss: 9.1421 - acc: 0.408 - ETA: 4s - loss: 9.1511 - acc: 0.407 - ETA: 4s - loss: 9.1527 - acc: 0.407 - ETA: 4s - loss: 9.1600 - acc: 0.406 - ETA: 4s - loss: 9.1756 - acc: 0.406 - ETA: 4s - loss: 9.1873 - acc: 0.405 - ETA: 4s - loss: 9.2007 - acc: 0.404 - ETA: 4s - loss: 9.2022 - acc: 0.404 - ETA: 4s - loss: 9.2082 - acc: 0.404 - ETA: 4s - loss: 9.2015 - acc: 0.404 - ETA: 4s - loss: 9.1915 - acc: 0.404 - ETA: 4s - loss: 9.1910 - acc: 0.404 - ETA: 4s - loss: 9.2019 - acc: 0.403 - ETA: 3s - loss: 9.2199 - acc: 0.402 - ETA: 3s - loss: 9.2229 - acc: 0.402 - ETA: 3s - loss: 9.2319 - acc: 0.401 - ETA: 3s - loss: 9.2400 - acc: 0.401 - ETA: 3s - loss: 9.2466 - acc: 0.400 - ETA: 3s - loss: 9.2474 - acc: 0.400 - ETA: 3s - loss: 9.2498 - acc: 0.400 - ETA: 3s - loss: 9.2605 - acc: 0.400 - ETA: 3s - loss: 9.2523 - acc: 0.400 - ETA: 3s - loss: 9.2396 - acc: 0.400 - ETA: 3s - loss: 9.2489 - acc: 0.400 - ETA: 3s - loss: 9.2617 - acc: 0.399 - ETA: 3s - loss: 9.2635 - acc: 0.399 - ETA: 3s - loss: 9.2832 - acc: 0.398 - ETA: 3s - loss: 9.2850 - acc: 0.398 - ETA: 3s - loss: 9.2735 - acc: 0.399 - ETA: 3s - loss: 9.2823 - acc: 0.398 - ETA: 3s - loss: 9.2947 - acc: 0.397 - ETA: 3s - loss: 9.2894 - acc: 0.397 - ETA: 3s - loss: 9.2964 - acc: 0.397 - ETA: 3s - loss: 9.2597 - acc: 0.399 - ETA: 3s - loss: 9.2590 - acc: 0.399 - ETA: 3s - loss: 9.2724 - acc: 0.399 - ETA: 2s - loss: 9.2654 - acc: 0.399 - ETA: 2s - loss: 9.2552 - acc: 0.399 - ETA: 2s - loss: 9.2750 - acc: 0.398 - ETA: 2s - loss: 9.2789 - acc: 0.398 - ETA: 2s - loss: 9.2950 - acc: 0.396 - ETA: 2s - loss: 9.2818 - acc: 0.397 - ETA: 2s - loss: 9.2749 - acc: 0.397 - ETA: 2s - loss: 9.2938 - acc: 0.396 - ETA: 2s - loss: 9.2969 - acc: 0.396 - ETA: 2s - loss: 9.3013 - acc: 0.396 - ETA: 2s - loss: 9.2818 - acc: 0.398 - ETA: 1s - loss: 9.2659 - acc: 0.399 - ETA: 1s - loss: 9.2696 - acc: 0.398 - ETA: 1s - loss: 9.2606 - acc: 0.399 - ETA: 1s - loss: 9.2517 - acc: 0.399 - ETA: 1s - loss: 9.2451 - acc: 0.400 - ETA: 1s - loss: 9.2397 - acc: 0.400 - ETA: 1s - loss: 9.2283 - acc: 0.401 - ETA: 1s - loss: 9.2440 - acc: 0.400 - ETA: 1s - loss: 9.2578 - acc: 0.399 - ETA: 1s - loss: 9.2584 - acc: 0.399 - ETA: 1s - loss: 9.2476 - acc: 0.400 - ETA: 1s - loss: 9.2409 - acc: 0.400 - ETA: 1s - loss: 9.2429 - acc: 0.400 - ETA: 0s - loss: 9.2526 - acc: 0.400 - ETA: 0s - loss: 9.2555 - acc: 0.399 - ETA: 0s - loss: 9.2495 - acc: 0.400 - ETA: 0s - loss: 9.2456 - acc: 0.400 - ETA: 0s - loss: 9.2522 - acc: 0.400 - ETA: 0s - loss: 9.2612 - acc: 0.399 - ETA: 0s - loss: 9.2660 - acc: 0.399 - ETA: 0s - loss: 9.2762 - acc: 0.398 - ETA: 0s - loss: 9.2855 - acc: 0.397 - ETA: 0s - loss: 9.2817 - acc: 0.398 - ETA: 0s - loss: 9.2884 - acc: 0.397 - ETA: 0s - loss: 9.2922 - acc: 0.397 - ETA: 0s - loss: 9.2899 - acc: 0.397 - ETA: 0s - loss: 9.2897 - acc: 0.397 - ETA: 0s - loss: 9.2785 - acc: 0.397 - ETA: 0s - loss: 9.2675 - acc: 0.398 - ETA: 0s - loss: 9.2663 - acc: 0.398 - ETA: 0s - loss: 9.2680 - acc: 0.398 - ETA: 0s - loss: 9.2753 - acc: 0.398 - ETA: 0s - loss: 9.2852 - acc: 0.397 - ETA: 0s - loss: 9.2836 - acc: 0.397 - ETA: 0s - loss: 9.2839 - acc: 0.397 - ETA: 0s - loss: 9.3057 - acc: 0.3961Epoch 00006: val_loss improved from 9.85014 to 9.53941, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 9.2949 - acc: 0.3966 - val_loss: 9.5394 - val_acc: 0.3509 Epoch 8/20 6640/6680 [============================>.] - ETA: 7s - loss: 8.0593 - acc: 0.500 - ETA: 5s - loss: 9.3643 - acc: 0.410 - ETA: 5s - loss: 8.9801 - acc: 0.437 - ETA: 5s - loss: 8.9703 - acc: 0.431 - ETA: 5s - loss: 8.7652 - acc: 0.435 - ETA: 6s - loss: 8.8317 - acc: 0.431 - ETA: 6s - loss: 9.0551 - acc: 0.418 - ETA: 6s - loss: 9.0500 - acc: 0.418 - ETA: 6s - loss: 9.0395 - acc: 0.416 - ETA: 6s - loss: 8.9688 - acc: 0.421 - ETA: 7s - loss: 8.9053 - acc: 0.425 - ETA: 7s - loss: 8.9514 - acc: 0.418 - ETA: 7s - loss: 8.9215 - acc: 0.421 - ETA: 7s - loss: 8.8954 - acc: 0.424 - ETA: 8s - loss: 8.9419 - acc: 0.422 - ETA: 8s - loss: 8.9173 - acc: 0.424 - ETA: 9s - loss: 8.8711 - acc: 0.427 - ETA: 10s - loss: 8.9147 - acc: 0.42 - ETA: 10s - loss: 8.8288 - acc: 0.43 - ETA: 11s - loss: 8.7969 - acc: 0.43 - ETA: 11s - loss: 8.7986 - acc: 0.43 - ETA: 11s - loss: 8.8396 - acc: 0.43 - ETA: 11s - loss: 8.7471 - acc: 0.43 - ETA: 10s - loss: 8.7584 - acc: 0.43 - ETA: 10s - loss: 8.6805 - acc: 0.44 - ETA: 10s - loss: 8.7525 - acc: 0.43 - ETA: 10s - loss: 8.7620 - acc: 0.43 - ETA: 9s - loss: 8.7470 - acc: 0.4384 - ETA: 9s - loss: 8.8079 - acc: 0.435 - ETA: 9s - loss: 8.7854 - acc: 0.435 - ETA: 9s - loss: 8.8288 - acc: 0.433 - ETA: 8s - loss: 8.8069 - acc: 0.433 - ETA: 8s - loss: 8.8559 - acc: 0.431 - ETA: 8s - loss: 8.9234 - acc: 0.427 - ETA: 8s - loss: 8.9109 - acc: 0.426 - ETA: 7s - loss: 8.9104 - acc: 0.425 - ETA: 7s - loss: 8.9560 - acc: 0.421 - ETA: 7s - loss: 8.9626 - acc: 0.421 - ETA: 7s - loss: 8.9848 - acc: 0.421 - ETA: 7s - loss: 9.0171 - acc: 0.419 - ETA: 7s - loss: 9.0286 - acc: 0.418 - ETA: 7s - loss: 8.9690 - acc: 0.422 - ETA: 6s - loss: 9.0172 - acc: 0.419 - ETA: 6s - loss: 8.9678 - acc: 0.422 - ETA: 6s - loss: 8.9837 - acc: 0.421 - ETA: 6s - loss: 9.0195 - acc: 0.419 - ETA: 6s - loss: 9.0435 - acc: 0.417 - ETA: 6s - loss: 9.0407 - acc: 0.418 - ETA: 6s - loss: 9.0556 - acc: 0.415 - ETA: 6s - loss: 9.0815 - acc: 0.414 - ETA: 6s - loss: 9.0905 - acc: 0.414 - ETA: 6s - loss: 9.0736 - acc: 0.415 - ETA: 6s - loss: 9.0342 - acc: 0.417 - ETA: 6s - loss: 9.0351 - acc: 0.416 - ETA: 6s - loss: 9.0492 - acc: 0.415 - ETA: 6s - loss: 9.0252 - acc: 0.417 - ETA: 6s - loss: 9.0384 - acc: 0.416 - ETA: 6s - loss: 9.0701 - acc: 0.414 - ETA: 6s - loss: 9.0562 - acc: 0.415 - ETA: 6s - loss: 9.0562 - acc: 0.414 - ETA: 6s - loss: 9.0680 - acc: 0.414 - ETA: 6s - loss: 9.0847 - acc: 0.413 - ETA: 6s - loss: 9.0717 - acc: 0.413 - ETA: 6s - loss: 9.0579 - acc: 0.413 - ETA: 6s - loss: 9.0489 - acc: 0.414 - ETA: 6s - loss: 9.0480 - acc: 0.413 - ETA: 6s - loss: 9.0598 - acc: 0.413 - ETA: 6s - loss: 9.0874 - acc: 0.411 - ETA: 6s - loss: 9.0700 - acc: 0.412 - ETA: 6s - loss: 9.0638 - acc: 0.412 - ETA: 6s - loss: 9.0462 - acc: 0.413 - ETA: 6s - loss: 9.0700 - acc: 0.411 - ETA: 5s - loss: 9.0298 - acc: 0.412 - ETA: 5s - loss: 9.0368 - acc: 0.412 - ETA: 5s - loss: 9.0010 - acc: 0.413 - ETA: 5s - loss: 9.0063 - acc: 0.413 - ETA: 5s - loss: 9.0083 - acc: 0.413 - ETA: 5s - loss: 9.0223 - acc: 0.412 - ETA: 5s - loss: 9.0281 - acc: 0.411 - ETA: 5s - loss: 9.0495 - acc: 0.410 - ETA: 5s - loss: 9.0386 - acc: 0.411 - ETA: 4s - loss: 9.0343 - acc: 0.410 - ETA: 4s - loss: 9.0192 - acc: 0.412 - ETA: 4s - loss: 9.0041 - acc: 0.413 - ETA: 4s - loss: 8.9825 - acc: 0.414 - ETA: 4s - loss: 8.9804 - acc: 0.414 - ETA: 4s - loss: 8.9830 - acc: 0.414 - ETA: 4s - loss: 8.9808 - acc: 0.414 - ETA: 4s - loss: 8.9610 - acc: 0.416 - ETA: 4s - loss: 8.9811 - acc: 0.414 - ETA: 4s - loss: 8.9760 - acc: 0.415 - ETA: 4s - loss: 8.9635 - acc: 0.416 - ETA: 4s - loss: 8.9696 - acc: 0.415 - ETA: 4s - loss: 8.9591 - acc: 0.415 - ETA: 3s - loss: 8.9306 - acc: 0.417 - ETA: 3s - loss: 8.9130 - acc: 0.418 - ETA: 3s - loss: 8.9283 - acc: 0.417 - ETA: 3s - loss: 8.9357 - acc: 0.416 - ETA: 3s - loss: 8.9371 - acc: 0.416 - ETA: 3s - loss: 8.9372 - acc: 0.416 - ETA: 3s - loss: 8.9481 - acc: 0.415 - ETA: 3s - loss: 8.9340 - acc: 0.416 - ETA: 3s - loss: 8.9308 - acc: 0.416 - ETA: 3s - loss: 8.9306 - acc: 0.416 - ETA: 3s - loss: 8.9305 - acc: 0.416 - ETA: 3s - loss: 8.9282 - acc: 0.416 - ETA: 3s - loss: 8.9292 - acc: 0.416 - ETA: 3s - loss: 8.9338 - acc: 0.416 - ETA: 3s - loss: 8.9481 - acc: 0.415 - ETA: 3s - loss: 8.9552 - acc: 0.415 - ETA: 3s - loss: 8.9520 - acc: 0.415 - ETA: 3s - loss: 8.9587 - acc: 0.415 - ETA: 3s - loss: 8.9477 - acc: 0.415 - ETA: 3s - loss: 8.9512 - acc: 0.415 - ETA: 3s - loss: 8.9548 - acc: 0.414 - ETA: 3s - loss: 8.9617 - acc: 0.414 - ETA: 3s - loss: 8.9507 - acc: 0.414 - ETA: 3s - loss: 8.9508 - acc: 0.414 - ETA: 2s - loss: 8.9564 - acc: 0.414 - ETA: 2s - loss: 8.9597 - acc: 0.414 - ETA: 2s - loss: 8.9690 - acc: 0.414 - ETA: 2s - loss: 8.9732 - acc: 0.413 - ETA: 2s - loss: 8.9774 - acc: 0.413 - ETA: 2s - loss: 8.9890 - acc: 0.412 - ETA: 2s - loss: 8.9981 - acc: 0.412 - ETA: 2s - loss: 8.9826 - acc: 0.413 - ETA: 2s - loss: 8.9770 - acc: 0.413 - ETA: 2s - loss: 8.9840 - acc: 0.413 - ETA: 2s - loss: 8.9925 - acc: 0.412 - ETA: 1s - loss: 8.9977 - acc: 0.412 - ETA: 1s - loss: 9.0068 - acc: 0.411 - ETA: 1s - loss: 9.0001 - acc: 0.412 - ETA: 1s - loss: 9.0034 - acc: 0.412 - ETA: 1s - loss: 8.9862 - acc: 0.413 - ETA: 1s - loss: 9.0061 - acc: 0.411 - ETA: 1s - loss: 9.0052 - acc: 0.412 - ETA: 1s - loss: 9.0177 - acc: 0.411 - ETA: 1s - loss: 9.0080 - acc: 0.411 - ETA: 1s - loss: 9.0000 - acc: 0.412 - ETA: 1s - loss: 9.0001 - acc: 0.412 - ETA: 1s - loss: 8.9992 - acc: 0.412 - ETA: 0s - loss: 8.9975 - acc: 0.412 - ETA: 0s - loss: 8.9940 - acc: 0.412 - ETA: 0s - loss: 8.9882 - acc: 0.412 - ETA: 0s - loss: 8.9805 - acc: 0.413 - ETA: 0s - loss: 8.9801 - acc: 0.413 - ETA: 0s - loss: 8.9875 - acc: 0.412 - ETA: 0s - loss: 8.9871 - acc: 0.412 - ETA: 0s - loss: 8.9849 - acc: 0.413 - ETA: 0s - loss: 8.9847 - acc: 0.413 - ETA: 0s - loss: 8.9971 - acc: 0.412 - ETA: 0s - loss: 8.9967 - acc: 0.412 - ETA: 0s - loss: 8.9989 - acc: 0.412 - ETA: 0s - loss: 9.0023 - acc: 0.412 - ETA: 0s - loss: 9.0018 - acc: 0.412 - ETA: 0s - loss: 8.9898 - acc: 0.412 - ETA: 0s - loss: 8.9898 - acc: 0.412 - ETA: 0s - loss: 8.9910 - acc: 0.412 - ETA: 0s - loss: 8.9917 - acc: 0.412 - ETA: 0s - loss: 8.9797 - acc: 0.413 - ETA: 0s - loss: 8.9804 - acc: 0.413 - ETA: 0s - loss: 8.9768 - acc: 0.413 - ETA: 0s - loss: 8.9690 - acc: 0.4137Epoch 00007: val_loss improved from 9.53941 to 9.22024, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 11s - loss: 8.9781 - acc: 0.4133 - val_loss: 9.2202 - val_acc: 0.3653 Epoch 9/20 6660/6680 [============================>.] - ETA: 8s - loss: 6.7482 - acc: 0.550 - ETA: 6s - loss: 8.3953 - acc: 0.462 - ETA: 6s - loss: 8.4519 - acc: 0.464 - ETA: 6s - loss: 8.3502 - acc: 0.465 - ETA: 6s - loss: 8.5343 - acc: 0.457 - ETA: 6s - loss: 8.7875 - acc: 0.440 - ETA: 6s - loss: 8.6924 - acc: 0.444 - ETA: 6s - loss: 8.7156 - acc: 0.442 - ETA: 6s - loss: 8.6006 - acc: 0.447 - ETA: 6s - loss: 8.7430 - acc: 0.438 - ETA: 7s - loss: 8.8390 - acc: 0.433 - ETA: 7s - loss: 8.8236 - acc: 0.431 - ETA: 7s - loss: 8.6264 - acc: 0.445 - ETA: 7s - loss: 8.6899 - acc: 0.441 - ETA: 8s - loss: 8.7383 - acc: 0.438 - ETA: 8s - loss: 8.6527 - acc: 0.444 - ETA: 9s - loss: 8.7246 - acc: 0.440 - ETA: 10s - loss: 8.7112 - acc: 0.44 - ETA: 10s - loss: 8.6961 - acc: 0.44 - ETA: 11s - loss: 8.7004 - acc: 0.44 - ETA: 11s - loss: 8.7243 - acc: 0.44 - ETA: 11s - loss: 8.6788 - acc: 0.44 - ETA: 11s - loss: 8.6800 - acc: 0.44 - ETA: 11s - loss: 8.6809 - acc: 0.44 - ETA: 10s - loss: 8.6311 - acc: 0.44 - ETA: 10s - loss: 8.6874 - acc: 0.44 - ETA: 10s - loss: 8.7406 - acc: 0.43 - ETA: 9s - loss: 8.7838 - acc: 0.4328 - ETA: 9s - loss: 8.7691 - acc: 0.431 - ETA: 9s - loss: 8.8093 - acc: 0.428 - ETA: 9s - loss: 8.8434 - acc: 0.426 - ETA: 8s - loss: 8.8205 - acc: 0.426 - ETA: 8s - loss: 8.7996 - acc: 0.427 - ETA: 8s - loss: 8.8661 - acc: 0.424 - ETA: 8s - loss: 8.8425 - acc: 0.423 - ETA: 8s - loss: 8.7857 - acc: 0.427 - ETA: 8s - loss: 8.7341 - acc: 0.429 - ETA: 7s - loss: 8.7329 - acc: 0.428 - ETA: 7s - loss: 8.7204 - acc: 0.429 - ETA: 7s - loss: 8.7249 - acc: 0.429 - ETA: 7s - loss: 8.7312 - acc: 0.429 - ETA: 7s - loss: 8.6770 - acc: 0.433 - ETA: 7s - loss: 8.6825 - acc: 0.433 - ETA: 7s - loss: 8.6856 - acc: 0.432 - ETA: 6s - loss: 8.7138 - acc: 0.431 - ETA: 6s - loss: 8.7167 - acc: 0.431 - ETA: 6s - loss: 8.6775 - acc: 0.434 - ETA: 6s - loss: 8.6928 - acc: 0.432 - ETA: 6s - loss: 8.7047 - acc: 0.432 - ETA: 6s - loss: 8.7146 - acc: 0.432 - ETA: 6s - loss: 8.7060 - acc: 0.432 - ETA: 6s - loss: 8.7167 - acc: 0.431 - ETA: 6s - loss: 8.6871 - acc: 0.433 - ETA: 6s - loss: 8.6835 - acc: 0.434 - ETA: 6s - loss: 8.6385 - acc: 0.437 - ETA: 6s - loss: 8.6365 - acc: 0.437 - ETA: 6s - loss: 8.6371 - acc: 0.437 - ETA: 6s - loss: 8.6392 - acc: 0.437 - ETA: 6s - loss: 8.6286 - acc: 0.438 - ETA: 6s - loss: 8.6556 - acc: 0.436 - ETA: 6s - loss: 8.6511 - acc: 0.437 - ETA: 6s - loss: 8.6656 - acc: 0.436 - ETA: 6s - loss: 8.6505 - acc: 0.436 - ETA: 6s - loss: 8.6524 - acc: 0.436 - ETA: 6s - loss: 8.6487 - acc: 0.436 - ETA: 6s - loss: 8.6188 - acc: 0.438 - ETA: 6s - loss: 8.6119 - acc: 0.438 - ETA: 6s - loss: 8.5951 - acc: 0.439 - ETA: 6s - loss: 8.6176 - acc: 0.438 - ETA: 6s - loss: 8.6237 - acc: 0.437 - ETA: 6s - loss: 8.6182 - acc: 0.437 - ETA: 6s - loss: 8.6293 - acc: 0.437 - ETA: 6s - loss: 8.6013 - acc: 0.438 - ETA: 5s - loss: 8.6078 - acc: 0.437 - ETA: 5s - loss: 8.6215 - acc: 0.437 - ETA: 5s - loss: 8.6183 - acc: 0.437 - ETA: 5s - loss: 8.6466 - acc: 0.435 - ETA: 5s - loss: 8.6418 - acc: 0.436 - ETA: 5s - loss: 8.6718 - acc: 0.434 - ETA: 5s - loss: 8.6642 - acc: 0.435 - ETA: 5s - loss: 8.6792 - acc: 0.434 - ETA: 5s - loss: 8.6959 - acc: 0.433 - ETA: 5s - loss: 8.7070 - acc: 0.432 - ETA: 4s - loss: 8.6884 - acc: 0.433 - ETA: 4s - loss: 8.6706 - acc: 0.434 - ETA: 4s - loss: 8.6828 - acc: 0.433 - ETA: 4s - loss: 8.6800 - acc: 0.433 - ETA: 4s - loss: 8.6579 - acc: 0.435 - ETA: 4s - loss: 8.6564 - acc: 0.434 - ETA: 4s - loss: 8.6689 - acc: 0.434 - ETA: 4s - loss: 8.6752 - acc: 0.433 - ETA: 4s - loss: 8.6738 - acc: 0.433 - ETA: 4s - loss: 8.6747 - acc: 0.433 - ETA: 4s - loss: 8.6805 - acc: 0.433 - ETA: 4s - loss: 8.6825 - acc: 0.433 - ETA: 3s - loss: 8.6833 - acc: 0.432 - ETA: 3s - loss: 8.6664 - acc: 0.434 - ETA: 3s - loss: 8.6721 - acc: 0.433 - ETA: 3s - loss: 8.6742 - acc: 0.433 - ETA: 3s - loss: 8.6761 - acc: 0.434 - ETA: 3s - loss: 8.6706 - acc: 0.434 - ETA: 3s - loss: 8.6580 - acc: 0.435 - ETA: 3s - loss: 8.6479 - acc: 0.435 - ETA: 3s - loss: 8.6499 - acc: 0.435 - ETA: 3s - loss: 8.6582 - acc: 0.435 - ETA: 3s - loss: 8.6537 - acc: 0.435 - ETA: 3s - loss: 8.6482 - acc: 0.435 - ETA: 3s - loss: 8.6527 - acc: 0.435 - ETA: 3s - loss: 8.6593 - acc: 0.435 - ETA: 3s - loss: 8.6712 - acc: 0.434 - ETA: 3s - loss: 8.6721 - acc: 0.434 - ETA: 3s - loss: 8.6620 - acc: 0.435 - ETA: 3s - loss: 8.6737 - acc: 0.434 - ETA: 3s - loss: 8.6887 - acc: 0.433 - ETA: 3s - loss: 8.6804 - acc: 0.433 - ETA: 3s - loss: 8.6720 - acc: 0.434 - ETA: 3s - loss: 8.6586 - acc: 0.434 - ETA: 3s - loss: 8.6473 - acc: 0.435 - ETA: 2s - loss: 8.6467 - acc: 0.435 - ETA: 2s - loss: 8.6424 - acc: 0.435 - ETA: 2s - loss: 8.6233 - acc: 0.437 - ETA: 2s - loss: 8.6356 - acc: 0.436 - ETA: 2s - loss: 8.6303 - acc: 0.436 - ETA: 2s - loss: 8.6398 - acc: 0.436 - ETA: 2s - loss: 8.6295 - acc: 0.436 - ETA: 2s - loss: 8.6309 - acc: 0.436 - ETA: 2s - loss: 8.6275 - acc: 0.437 - ETA: 2s - loss: 8.6383 - acc: 0.436 - ETA: 2s - loss: 8.6388 - acc: 0.436 - ETA: 2s - loss: 8.6393 - acc: 0.436 - ETA: 1s - loss: 8.6452 - acc: 0.436 - ETA: 1s - loss: 8.6497 - acc: 0.436 - ETA: 1s - loss: 8.6568 - acc: 0.435 - ETA: 1s - loss: 8.6670 - acc: 0.435 - ETA: 1s - loss: 8.6602 - acc: 0.435 - ETA: 1s - loss: 8.6572 - acc: 0.435 - ETA: 1s - loss: 8.6541 - acc: 0.435 - ETA: 1s - loss: 8.6534 - acc: 0.436 - ETA: 1s - loss: 8.6618 - acc: 0.435 - ETA: 1s - loss: 8.6577 - acc: 0.435 - ETA: 1s - loss: 8.6511 - acc: 0.436 - ETA: 1s - loss: 8.6554 - acc: 0.436 - ETA: 1s - loss: 8.6615 - acc: 0.436 - ETA: 1s - loss: 8.6744 - acc: 0.435 - ETA: 0s - loss: 8.6824 - acc: 0.434 - ETA: 0s - loss: 8.6887 - acc: 0.434 - ETA: 0s - loss: 8.6795 - acc: 0.435 - ETA: 0s - loss: 8.6749 - acc: 0.435 - ETA: 0s - loss: 8.6703 - acc: 0.435 - ETA: 0s - loss: 8.6660 - acc: 0.436 - ETA: 0s - loss: 8.6596 - acc: 0.436 - ETA: 0s - loss: 8.6525 - acc: 0.436 - ETA: 0s - loss: 8.6520 - acc: 0.436 - ETA: 0s - loss: 8.6476 - acc: 0.436 - ETA: 0s - loss: 8.6462 - acc: 0.436 - ETA: 0s - loss: 8.6417 - acc: 0.437 - ETA: 0s - loss: 8.6374 - acc: 0.437 - ETA: 0s - loss: 8.6388 - acc: 0.437 - ETA: 0s - loss: 8.6350 - acc: 0.437 - ETA: 0s - loss: 8.6309 - acc: 0.437 - ETA: 0s - loss: 8.6267 - acc: 0.438 - ETA: 0s - loss: 8.6277 - acc: 0.437 - ETA: 0s - loss: 8.6103 - acc: 0.438 - ETA: 0s - loss: 8.6070 - acc: 0.439 - ETA: 0s - loss: 8.6091 - acc: 0.438 - ETA: 0s - loss: 8.6058 - acc: 0.439 - ETA: 0s - loss: 8.6054 - acc: 0.4392Epoch 00008: val_loss improved from 9.22024 to 8.97945, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 8.6110 - acc: 0.4389 - val_loss: 8.9794 - val_acc: 0.3796 Epoch 10/20 6640/6680 [============================>.] - ETA: 8s - loss: 9.6709 - acc: 0.400 - ETA: 6s - loss: 9.0669 - acc: 0.437 - ETA: 6s - loss: 8.7504 - acc: 0.457 - ETA: 7s - loss: 8.9603 - acc: 0.438 - ETA: 7s - loss: 9.4733 - acc: 0.404 - ETA: 7s - loss: 9.2195 - acc: 0.414 - ETA: 7s - loss: 9.0629 - acc: 0.426 - ETA: 7s - loss: 8.9774 - acc: 0.431 - ETA: 7s - loss: 8.9296 - acc: 0.435 - ETA: 7s - loss: 9.0095 - acc: 0.430 - ETA: 8s - loss: 8.9122 - acc: 0.433 - ETA: 8s - loss: 8.9107 - acc: 0.434 - ETA: 9s - loss: 8.8553 - acc: 0.436 - ETA: 9s - loss: 8.9154 - acc: 0.433 - ETA: 10s - loss: 8.9136 - acc: 0.43 - ETA: 10s - loss: 8.8044 - acc: 0.43 - ETA: 11s - loss: 8.8870 - acc: 0.43 - ETA: 12s - loss: 8.8174 - acc: 0.43 - ETA: 12s - loss: 8.8442 - acc: 0.43 - ETA: 13s - loss: 8.8448 - acc: 0.43 - ETA: 12s - loss: 8.8692 - acc: 0.43 - ETA: 12s - loss: 8.8704 - acc: 0.43 - ETA: 11s - loss: 8.8127 - acc: 0.44 - ETA: 11s - loss: 8.7799 - acc: 0.44 - ETA: 10s - loss: 8.8362 - acc: 0.44 - ETA: 10s - loss: 8.8052 - acc: 0.44 - ETA: 10s - loss: 8.6910 - acc: 0.44 - ETA: 9s - loss: 8.5898 - acc: 0.4545 - ETA: 9s - loss: 8.5765 - acc: 0.456 - ETA: 9s - loss: 8.6441 - acc: 0.452 - ETA: 8s - loss: 8.6301 - acc: 0.453 - ETA: 8s - loss: 8.5709 - acc: 0.457 - ETA: 8s - loss: 8.5244 - acc: 0.460 - ETA: 8s - loss: 8.5499 - acc: 0.459 - ETA: 8s - loss: 8.5102 - acc: 0.461 - ETA: 8s - loss: 8.5032 - acc: 0.462 - ETA: 7s - loss: 8.4315 - acc: 0.466 - ETA: 7s - loss: 8.4515 - acc: 0.465 - ETA: 7s - loss: 8.4615 - acc: 0.464 - ETA: 7s - loss: 8.4702 - acc: 0.464 - ETA: 7s - loss: 8.5112 - acc: 0.460 - ETA: 7s - loss: 8.4819 - acc: 0.462 - ETA: 6s - loss: 8.4624 - acc: 0.463 - ETA: 6s - loss: 8.4866 - acc: 0.462 - ETA: 6s - loss: 8.4784 - acc: 0.463 - ETA: 6s - loss: 8.4319 - acc: 0.466 - ETA: 6s - loss: 8.4174 - acc: 0.467 - ETA: 6s - loss: 8.3990 - acc: 0.468 - ETA: 6s - loss: 8.3893 - acc: 0.469 - ETA: 6s - loss: 8.3616 - acc: 0.471 - ETA: 6s - loss: 8.3716 - acc: 0.470 - ETA: 6s - loss: 8.3526 - acc: 0.471 - ETA: 6s - loss: 8.3896 - acc: 0.469 - ETA: 6s - loss: 8.4325 - acc: 0.466 - ETA: 6s - loss: 8.3959 - acc: 0.468 - ETA: 6s - loss: 8.4135 - acc: 0.467 - ETA: 6s - loss: 8.3974 - acc: 0.468 - ETA: 6s - loss: 8.4145 - acc: 0.467 - ETA: 6s - loss: 8.3987 - acc: 0.468 - ETA: 6s - loss: 8.3776 - acc: 0.469 - ETA: 6s - loss: 8.3709 - acc: 0.469 - ETA: 6s - loss: 8.3969 - acc: 0.468 - ETA: 6s - loss: 8.4261 - acc: 0.466 - ETA: 6s - loss: 8.4147 - acc: 0.466 - ETA: 7s - loss: 8.4082 - acc: 0.466 - ETA: 7s - loss: 8.3934 - acc: 0.467 - ETA: 7s - loss: 8.4214 - acc: 0.465 - ETA: 6s - loss: 8.4521 - acc: 0.463 - ETA: 6s - loss: 8.4648 - acc: 0.462 - ETA: 6s - loss: 8.4596 - acc: 0.462 - ETA: 6s - loss: 8.4722 - acc: 0.461 - ETA: 6s - loss: 8.4673 - acc: 0.461 - ETA: 6s - loss: 8.4685 - acc: 0.461 - ETA: 6s - loss: 8.4627 - acc: 0.460 - ETA: 6s - loss: 8.4357 - acc: 0.462 - ETA: 6s - loss: 8.4169 - acc: 0.463 - ETA: 6s - loss: 8.4076 - acc: 0.463 - ETA: 5s - loss: 8.3945 - acc: 0.464 - ETA: 5s - loss: 8.4260 - acc: 0.462 - ETA: 5s - loss: 8.4144 - acc: 0.463 - ETA: 5s - loss: 8.4276 - acc: 0.462 - ETA: 5s - loss: 8.4212 - acc: 0.463 - ETA: 5s - loss: 8.4277 - acc: 0.462 - ETA: 5s - loss: 8.4495 - acc: 0.461 - ETA: 5s - loss: 8.4635 - acc: 0.461 - ETA: 5s - loss: 8.4500 - acc: 0.462 - ETA: 5s - loss: 8.4465 - acc: 0.462 - ETA: 4s - loss: 8.4690 - acc: 0.460 - ETA: 4s - loss: 8.4728 - acc: 0.459 - ETA: 4s - loss: 8.4733 - acc: 0.459 - ETA: 4s - loss: 8.4477 - acc: 0.461 - ETA: 4s - loss: 8.4755 - acc: 0.459 - ETA: 4s - loss: 8.4921 - acc: 0.458 - ETA: 4s - loss: 8.5082 - acc: 0.457 - ETA: 4s - loss: 8.5281 - acc: 0.456 - ETA: 4s - loss: 8.5118 - acc: 0.457 - ETA: 4s - loss: 8.5211 - acc: 0.457 - ETA: 4s - loss: 8.5169 - acc: 0.457 - ETA: 4s - loss: 8.5346 - acc: 0.456 - ETA: 4s - loss: 8.5612 - acc: 0.454 - ETA: 4s - loss: 8.5549 - acc: 0.455 - ETA: 4s - loss: 8.5565 - acc: 0.455 - ETA: 4s - loss: 8.5503 - acc: 0.455 - ETA: 4s - loss: 8.5566 - acc: 0.454 - ETA: 4s - loss: 8.5657 - acc: 0.454 - ETA: 4s - loss: 8.5596 - acc: 0.454 - ETA: 4s - loss: 8.5588 - acc: 0.454 - ETA: 4s - loss: 8.5565 - acc: 0.454 - ETA: 4s - loss: 8.5640 - acc: 0.454 - ETA: 4s - loss: 8.5729 - acc: 0.453 - ETA: 3s - loss: 8.5668 - acc: 0.454 - ETA: 3s - loss: 8.5769 - acc: 0.453 - ETA: 3s - loss: 8.5725 - acc: 0.453 - ETA: 3s - loss: 8.5674 - acc: 0.454 - ETA: 3s - loss: 8.5669 - acc: 0.454 - ETA: 3s - loss: 8.5661 - acc: 0.454 - ETA: 3s - loss: 8.5552 - acc: 0.454 - ETA: 3s - loss: 8.5441 - acc: 0.455 - ETA: 3s - loss: 8.5480 - acc: 0.455 - ETA: 3s - loss: 8.5549 - acc: 0.454 - ETA: 3s - loss: 8.5448 - acc: 0.455 - ETA: 3s - loss: 8.5355 - acc: 0.455 - ETA: 3s - loss: 8.5420 - acc: 0.455 - ETA: 2s - loss: 8.5228 - acc: 0.456 - ETA: 2s - loss: 8.5354 - acc: 0.455 - ETA: 2s - loss: 8.5161 - acc: 0.456 - ETA: 2s - loss: 8.5030 - acc: 0.457 - ETA: 2s - loss: 8.4963 - acc: 0.458 - ETA: 2s - loss: 8.5025 - acc: 0.457 - ETA: 2s - loss: 8.4990 - acc: 0.458 - ETA: 2s - loss: 8.4972 - acc: 0.458 - ETA: 2s - loss: 8.4944 - acc: 0.458 - ETA: 2s - loss: 8.4977 - acc: 0.458 - ETA: 2s - loss: 8.5069 - acc: 0.457 - ETA: 2s - loss: 8.5279 - acc: 0.456 - ETA: 2s - loss: 8.5244 - acc: 0.456 - ETA: 2s - loss: 8.5287 - acc: 0.455 - ETA: 1s - loss: 8.5322 - acc: 0.455 - ETA: 1s - loss: 8.5446 - acc: 0.454 - ETA: 1s - loss: 8.5430 - acc: 0.454 - ETA: 1s - loss: 8.5425 - acc: 0.455 - ETA: 1s - loss: 8.5421 - acc: 0.455 - ETA: 1s - loss: 8.5432 - acc: 0.454 - ETA: 1s - loss: 8.5408 - acc: 0.454 - ETA: 1s - loss: 8.5459 - acc: 0.454 - ETA: 1s - loss: 8.5459 - acc: 0.454 - ETA: 1s - loss: 8.5344 - acc: 0.455 - ETA: 1s - loss: 8.5316 - acc: 0.455 - ETA: 1s - loss: 8.5448 - acc: 0.454 - ETA: 1s - loss: 8.5451 - acc: 0.454 - ETA: 1s - loss: 8.5517 - acc: 0.454 - ETA: 1s - loss: 8.5527 - acc: 0.454 - ETA: 1s - loss: 8.5646 - acc: 0.453 - ETA: 1s - loss: 8.5577 - acc: 0.453 - ETA: 1s - loss: 8.5606 - acc: 0.453 - ETA: 1s - loss: 8.5509 - acc: 0.454 - ETA: 1s - loss: 8.5466 - acc: 0.454 - ETA: 1s - loss: 8.5503 - acc: 0.454 - ETA: 1s - loss: 8.5567 - acc: 0.454 - ETA: 1s - loss: 8.5498 - acc: 0.454 - ETA: 0s - loss: 8.5534 - acc: 0.454 - ETA: 0s - loss: 8.5609 - acc: 0.453 - ETA: 0s - loss: 8.5590 - acc: 0.453 - ETA: 0s - loss: 8.5507 - acc: 0.454 - ETA: 0s - loss: 8.5373 - acc: 0.455 - ETA: 0s - loss: 8.5400 - acc: 0.454 - ETA: 0s - loss: 8.5422 - acc: 0.454 - ETA: 0s - loss: 8.5375 - acc: 0.455 - ETA: 0s - loss: 8.5370 - acc: 0.455 - ETA: 0s - loss: 8.5294 - acc: 0.455 - ETA: 0s - loss: 8.5318 - acc: 0.455 - ETA: 0s - loss: 8.5245 - acc: 0.455 - ETA: 0s - loss: 8.5268 - acc: 0.455 - ETA: 0s - loss: 8.5318 - acc: 0.4553Epoch 00009: val_loss did not improve 6680/6680 [==============================] - 12s - loss: 8.5130 - acc: 0.4563 - val_loss: 9.0172 - val_acc: 0.3952 Epoch 11/20 6640/6680 [============================>.] - ETA: 8s - loss: 7.3076 - acc: 0.550 - ETA: 7s - loss: 8.2764 - acc: 0.487 - ETA: 7s - loss: 8.5908 - acc: 0.457 - ETA: 7s - loss: 8.3706 - acc: 0.470 - ETA: 8s - loss: 8.1857 - acc: 0.483 - ETA: 7s - loss: 8.4310 - acc: 0.470 - ETA: 8s - loss: 8.4348 - acc: 0.470 - ETA: 9s - loss: 8.4139 - acc: 0.472 - ETA: 9s - loss: 8.6073 - acc: 0.460 - ETA: 9s - loss: 8.7009 - acc: 0.455 - ETA: 10s - loss: 8.7089 - acc: 0.45 - ETA: 11s - loss: 8.8017 - acc: 0.44 - ETA: 11s - loss: 8.7706 - acc: 0.45 - ETA: 11s - loss: 8.8086 - acc: 0.44 - ETA: 12s - loss: 8.6819 - acc: 0.45 - ETA: 12s - loss: 8.7200 - acc: 0.45 - ETA: 13s - loss: 8.6082 - acc: 0.45 - ETA: 13s - loss: 8.5310 - acc: 0.46 - ETA: 13s - loss: 8.3921 - acc: 0.47 - ETA: 12s - loss: 8.3308 - acc: 0.47 - ETA: 12s - loss: 8.3367 - acc: 0.47 - ETA: 11s - loss: 8.3151 - acc: 0.47 - ETA: 11s - loss: 8.1664 - acc: 0.48 - ETA: 10s - loss: 8.0902 - acc: 0.48 - ETA: 10s - loss: 8.1130 - acc: 0.48 - ETA: 9s - loss: 8.0303 - acc: 0.4920 - ETA: 9s - loss: 8.0316 - acc: 0.492 - ETA: 9s - loss: 8.0443 - acc: 0.491 - ETA: 9s - loss: 8.1177 - acc: 0.487 - ETA: 8s - loss: 8.1810 - acc: 0.483 - ETA: 8s - loss: 8.1881 - acc: 0.482 - ETA: 8s - loss: 8.1960 - acc: 0.482 - ETA: 8s - loss: 8.2238 - acc: 0.480 - ETA: 8s - loss: 8.2613 - acc: 0.478 - ETA: 7s - loss: 8.3385 - acc: 0.473 - ETA: 7s - loss: 8.3493 - acc: 0.472 - ETA: 7s - loss: 8.3783 - acc: 0.471 - ETA: 7s - loss: 8.4210 - acc: 0.468 - ETA: 7s - loss: 8.4498 - acc: 0.467 - ETA: 7s - loss: 8.4530 - acc: 0.466 - ETA: 7s - loss: 8.5004 - acc: 0.463 - ETA: 7s - loss: 8.4886 - acc: 0.463 - ETA: 6s - loss: 8.5136 - acc: 0.462 - ETA: 6s - loss: 8.4512 - acc: 0.465 - ETA: 6s - loss: 8.4629 - acc: 0.465 - ETA: 6s - loss: 8.4865 - acc: 0.463 - ETA: 6s - loss: 8.5014 - acc: 0.462 - ETA: 6s - loss: 8.4747 - acc: 0.464 - ETA: 6s - loss: 8.4360 - acc: 0.466 - ETA: 6s - loss: 8.4112 - acc: 0.467 - ETA: 6s - loss: 8.4081 - acc: 0.467 - ETA: 6s - loss: 8.4163 - acc: 0.467 - ETA: 6s - loss: 8.4630 - acc: 0.464 - ETA: 6s - loss: 8.4546 - acc: 0.464 - ETA: 6s - loss: 8.4584 - acc: 0.464 - ETA: 6s - loss: 8.4829 - acc: 0.462 - ETA: 6s - loss: 8.5066 - acc: 0.461 - ETA: 6s - loss: 8.5213 - acc: 0.460 - ETA: 6s - loss: 8.5309 - acc: 0.459 - ETA: 6s - loss: 8.5204 - acc: 0.460 - ETA: 7s - loss: 8.5366 - acc: 0.459 - ETA: 7s - loss: 8.5458 - acc: 0.458 - ETA: 7s - loss: 8.5484 - acc: 0.458 - ETA: 7s - loss: 8.5446 - acc: 0.459 - ETA: 7s - loss: 8.5411 - acc: 0.459 - ETA: 7s - loss: 8.5400 - acc: 0.459 - ETA: 7s - loss: 8.5513 - acc: 0.458 - ETA: 7s - loss: 8.5143 - acc: 0.460 - ETA: 6s - loss: 8.5118 - acc: 0.460 - ETA: 6s - loss: 8.5056 - acc: 0.460 - ETA: 6s - loss: 8.5227 - acc: 0.459 - ETA: 6s - loss: 8.4929 - acc: 0.461 - ETA: 6s - loss: 8.4540 - acc: 0.463 - ETA: 6s - loss: 8.4296 - acc: 0.465 - ETA: 6s - loss: 8.4245 - acc: 0.466 - ETA: 6s - loss: 8.4309 - acc: 0.465 - ETA: 6s - loss: 8.4100 - acc: 0.467 - ETA: 6s - loss: 8.4054 - acc: 0.467 - ETA: 5s - loss: 8.3801 - acc: 0.469 - ETA: 5s - loss: 8.3813 - acc: 0.469 - ETA: 5s - loss: 8.3661 - acc: 0.470 - ETA: 5s - loss: 8.3924 - acc: 0.468 - ETA: 5s - loss: 8.4258 - acc: 0.466 - ETA: 5s - loss: 8.4236 - acc: 0.466 - ETA: 5s - loss: 8.4463 - acc: 0.464 - ETA: 5s - loss: 8.4489 - acc: 0.464 - ETA: 5s - loss: 8.4584 - acc: 0.463 - ETA: 5s - loss: 8.4699 - acc: 0.462 - ETA: 5s - loss: 8.4852 - acc: 0.461 - ETA: 4s - loss: 8.4827 - acc: 0.461 - ETA: 4s - loss: 8.5021 - acc: 0.460 - ETA: 4s - loss: 8.4944 - acc: 0.460 - ETA: 4s - loss: 8.4582 - acc: 0.463 - ETA: 4s - loss: 8.4561 - acc: 0.463 - ETA: 4s - loss: 8.4729 - acc: 0.462 - ETA: 4s - loss: 8.4646 - acc: 0.462 - ETA: 4s - loss: 8.4694 - acc: 0.462 - ETA: 4s - loss: 8.4861 - acc: 0.461 - ETA: 4s - loss: 8.4886 - acc: 0.461 - ETA: 4s - loss: 8.4718 - acc: 0.461 - ETA: 4s - loss: 8.4658 - acc: 0.462 - ETA: 4s - loss: 8.4482 - acc: 0.463 - ETA: 4s - loss: 8.4385 - acc: 0.464 - ETA: 4s - loss: 8.4265 - acc: 0.464 - ETA: 4s - loss: 8.4192 - acc: 0.465 - ETA: 4s - loss: 8.4143 - acc: 0.465 - ETA: 4s - loss: 8.4095 - acc: 0.465 - ETA: 4s - loss: 8.4040 - acc: 0.466 - ETA: 4s - loss: 8.4060 - acc: 0.465 - ETA: 4s - loss: 8.4210 - acc: 0.464 - ETA: 4s - loss: 8.4015 - acc: 0.466 - ETA: 4s - loss: 8.3891 - acc: 0.466 - ETA: 3s - loss: 8.3937 - acc: 0.466 - ETA: 3s - loss: 8.3915 - acc: 0.466 - ETA: 3s - loss: 8.3835 - acc: 0.466 - ETA: 3s - loss: 8.3983 - acc: 0.465 - ETA: 3s - loss: 8.3922 - acc: 0.465 - ETA: 3s - loss: 8.4109 - acc: 0.464 - ETA: 3s - loss: 8.4149 - acc: 0.464 - ETA: 3s - loss: 8.4149 - acc: 0.464 - ETA: 3s - loss: 8.4163 - acc: 0.463 - ETA: 3s - loss: 8.4169 - acc: 0.463 - ETA: 3s - loss: 8.4173 - acc: 0.463 - ETA: 3s - loss: 8.4347 - acc: 0.462 - ETA: 3s - loss: 8.4307 - acc: 0.462 - ETA: 2s - loss: 8.4565 - acc: 0.460 - ETA: 2s - loss: 8.4502 - acc: 0.461 - ETA: 2s - loss: 8.4505 - acc: 0.461 - ETA: 2s - loss: 8.4545 - acc: 0.460 - ETA: 2s - loss: 8.4509 - acc: 0.460 - ETA: 2s - loss: 8.4511 - acc: 0.460 - ETA: 2s - loss: 8.4419 - acc: 0.461 - ETA: 2s - loss: 8.4381 - acc: 0.461 - ETA: 2s - loss: 8.4352 - acc: 0.462 - ETA: 2s - loss: 8.4477 - acc: 0.461 - ETA: 2s - loss: 8.4539 - acc: 0.461 - ETA: 2s - loss: 8.4630 - acc: 0.460 - ETA: 2s - loss: 8.4510 - acc: 0.461 - ETA: 2s - loss: 8.4367 - acc: 0.462 - ETA: 1s - loss: 8.4427 - acc: 0.461 - ETA: 1s - loss: 8.4384 - acc: 0.461 - ETA: 1s - loss: 8.4329 - acc: 0.462 - ETA: 1s - loss: 8.4291 - acc: 0.462 - ETA: 1s - loss: 8.4097 - acc: 0.463 - ETA: 1s - loss: 8.4064 - acc: 0.463 - ETA: 1s - loss: 8.4194 - acc: 0.462 - ETA: 1s - loss: 8.4256 - acc: 0.461 - ETA: 1s - loss: 8.4190 - acc: 0.462 - ETA: 1s - loss: 8.4290 - acc: 0.461 - ETA: 1s - loss: 8.4222 - acc: 0.462 - ETA: 1s - loss: 8.4126 - acc: 0.462 - ETA: 1s - loss: 8.4152 - acc: 0.462 - ETA: 1s - loss: 8.4223 - acc: 0.462 - ETA: 1s - loss: 8.4266 - acc: 0.461 - ETA: 1s - loss: 8.4156 - acc: 0.462 - ETA: 1s - loss: 8.4070 - acc: 0.462 - ETA: 1s - loss: 8.4076 - acc: 0.462 - ETA: 1s - loss: 8.4030 - acc: 0.462 - ETA: 1s - loss: 8.4035 - acc: 0.462 - ETA: 1s - loss: 8.4176 - acc: 0.461 - ETA: 1s - loss: 8.4081 - acc: 0.462 - ETA: 0s - loss: 8.4218 - acc: 0.461 - ETA: 0s - loss: 8.4248 - acc: 0.461 - ETA: 0s - loss: 8.4204 - acc: 0.461 - ETA: 0s - loss: 8.4207 - acc: 0.461 - ETA: 0s - loss: 8.4160 - acc: 0.461 - ETA: 0s - loss: 8.4123 - acc: 0.462 - ETA: 0s - loss: 8.4127 - acc: 0.462 - ETA: 0s - loss: 8.4039 - acc: 0.462 - ETA: 0s - loss: 8.4221 - acc: 0.461 - ETA: 0s - loss: 8.4155 - acc: 0.461 - ETA: 0s - loss: 8.4138 - acc: 0.461 - ETA: 0s - loss: 8.4117 - acc: 0.461 - ETA: 0s - loss: 8.4170 - acc: 0.461 - ETA: 0s - loss: 8.4232 - acc: 0.4607Epoch 00010: val_loss improved from 8.97945 to 8.79615, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 8.4139 - acc: 0.4614 - val_loss: 8.7962 - val_acc: 0.4024 Epoch 12/20 6660/6680 [============================>.] - ETA: 21s - loss: 7.4008 - acc: 0.50 - ETA: 20s - loss: 7.0567 - acc: 0.52 - ETA: 19s - loss: 7.9327 - acc: 0.48 - ETA: 19s - loss: 7.9644 - acc: 0.48 - ETA: 20s - loss: 7.5005 - acc: 0.52 - ETA: 21s - loss: 7.3261 - acc: 0.53 - ETA: 20s - loss: 7.3162 - acc: 0.53 - ETA: 21s - loss: 7.8134 - acc: 0.50 - ETA: 21s - loss: 8.0222 - acc: 0.49 - ETA: 21s - loss: 7.9568 - acc: 0.49 - ETA: 21s - loss: 8.0545 - acc: 0.48 - ETA: 19s - loss: 7.8693 - acc: 0.50 - ETA: 18s - loss: 7.9868 - acc: 0.49 - ETA: 16s - loss: 7.9481 - acc: 0.49 - ETA: 15s - loss: 8.2075 - acc: 0.48 - ETA: 14s - loss: 8.2993 - acc: 0.47 - ETA: 13s - loss: 8.3041 - acc: 0.47 - ETA: 13s - loss: 8.1390 - acc: 0.48 - ETA: 12s - loss: 8.2728 - acc: 0.47 - ETA: 12s - loss: 8.2933 - acc: 0.47 - ETA: 11s - loss: 8.3696 - acc: 0.47 - ETA: 11s - loss: 8.5029 - acc: 0.46 - ETA: 10s - loss: 8.3692 - acc: 0.47 - ETA: 10s - loss: 8.3160 - acc: 0.47 - ETA: 10s - loss: 8.2861 - acc: 0.47 - ETA: 10s - loss: 8.2314 - acc: 0.47 - ETA: 10s - loss: 8.2079 - acc: 0.48 - ETA: 9s - loss: 8.1719 - acc: 0.4830 - ETA: 9s - loss: 8.1842 - acc: 0.481 - ETA: 9s - loss: 8.1379 - acc: 0.484 - ETA: 9s - loss: 8.1785 - acc: 0.482 - ETA: 10s - loss: 8.2330 - acc: 0.47 - ETA: 10s - loss: 8.2439 - acc: 0.47 - ETA: 10s - loss: 8.2681 - acc: 0.47 - ETA: 10s - loss: 8.2517 - acc: 0.47 - ETA: 10s - loss: 8.2845 - acc: 0.47 - ETA: 10s - loss: 8.2526 - acc: 0.47 - ETA: 9s - loss: 8.2712 - acc: 0.4773 - ETA: 9s - loss: 8.2648 - acc: 0.477 - ETA: 9s - loss: 8.2789 - acc: 0.477 - ETA: 9s - loss: 8.3182 - acc: 0.474 - ETA: 9s - loss: 8.3511 - acc: 0.472 - ETA: 8s - loss: 8.3306 - acc: 0.473 - ETA: 8s - loss: 8.2862 - acc: 0.475 - ETA: 8s - loss: 8.3045 - acc: 0.474 - ETA: 8s - loss: 8.2964 - acc: 0.475 - ETA: 8s - loss: 8.2621 - acc: 0.476 - ETA: 8s - loss: 8.2845 - acc: 0.475 - ETA: 8s - loss: 8.3059 - acc: 0.473 - ETA: 8s - loss: 8.3093 - acc: 0.473 - ETA: 7s - loss: 8.2293 - acc: 0.479 - ETA: 7s - loss: 8.2176 - acc: 0.480 - ETA: 7s - loss: 8.2306 - acc: 0.479 - ETA: 7s - loss: 8.2260 - acc: 0.480 - ETA: 7s - loss: 8.2411 - acc: 0.479 - ETA: 7s - loss: 8.2302 - acc: 0.479 - ETA: 7s - loss: 8.2493 - acc: 0.478 - ETA: 7s - loss: 8.2534 - acc: 0.478 - ETA: 7s - loss: 8.2144 - acc: 0.481 - ETA: 7s - loss: 8.2060 - acc: 0.482 - ETA: 7s - loss: 8.2314 - acc: 0.480 - ETA: 7s - loss: 8.2560 - acc: 0.479 - ETA: 7s - loss: 8.2732 - acc: 0.478 - ETA: 7s - loss: 8.2878 - acc: 0.477 - ETA: 7s - loss: 8.2926 - acc: 0.477 - ETA: 7s - loss: 8.2983 - acc: 0.476 - ETA: 7s - loss: 8.2883 - acc: 0.476 - ETA: 7s - loss: 8.2807 - acc: 0.476 - ETA: 7s - loss: 8.2918 - acc: 0.476 - ETA: 7s - loss: 8.3034 - acc: 0.474 - ETA: 7s - loss: 8.3330 - acc: 0.473 - ETA: 7s - loss: 8.3308 - acc: 0.473 - ETA: 7s - loss: 8.3226 - acc: 0.473 - ETA: 7s - loss: 8.3255 - acc: 0.473 - ETA: 7s - loss: 8.2975 - acc: 0.475 - ETA: 7s - loss: 8.2595 - acc: 0.477 - ETA: 6s - loss: 8.2509 - acc: 0.478 - ETA: 6s - loss: 8.2377 - acc: 0.478 - ETA: 6s - loss: 8.2363 - acc: 0.478 - ETA: 6s - loss: 8.2251 - acc: 0.479 - ETA: 6s - loss: 8.2456 - acc: 0.477 - ETA: 6s - loss: 8.2438 - acc: 0.476 - ETA: 6s - loss: 8.2669 - acc: 0.475 - ETA: 6s - loss: 8.2955 - acc: 0.473 - ETA: 6s - loss: 8.2875 - acc: 0.474 - ETA: 5s - loss: 8.3067 - acc: 0.472 - ETA: 5s - loss: 8.2891 - acc: 0.473 - ETA: 5s - loss: 8.2765 - acc: 0.474 - ETA: 5s - loss: 8.2670 - acc: 0.475 - ETA: 5s - loss: 8.2611 - acc: 0.475 - ETA: 5s - loss: 8.2649 - acc: 0.475 - ETA: 5s - loss: 8.2677 - acc: 0.474 - ETA: 5s - loss: 8.2790 - acc: 0.474 - ETA: 5s - loss: 8.2938 - acc: 0.473 - ETA: 5s - loss: 8.2833 - acc: 0.473 - ETA: 5s - loss: 8.2729 - acc: 0.474 - ETA: 5s - loss: 8.2853 - acc: 0.473 - ETA: 5s - loss: 8.2841 - acc: 0.474 - ETA: 5s - loss: 8.2641 - acc: 0.475 - ETA: 5s - loss: 8.2615 - acc: 0.475 - ETA: 5s - loss: 8.2767 - acc: 0.474 - ETA: 4s - loss: 8.2897 - acc: 0.473 - ETA: 4s - loss: 8.3002 - acc: 0.473 - ETA: 4s - loss: 8.2803 - acc: 0.474 - ETA: 4s - loss: 8.2724 - acc: 0.474 - ETA: 4s - loss: 8.2765 - acc: 0.474 - ETA: 4s - loss: 8.2576 - acc: 0.475 - ETA: 4s - loss: 8.2653 - acc: 0.474 - ETA: 4s - loss: 8.2765 - acc: 0.473 - ETA: 4s - loss: 8.2707 - acc: 0.473 - ETA: 4s - loss: 8.2553 - acc: 0.474 - ETA: 4s - loss: 8.2491 - acc: 0.475 - ETA: 4s - loss: 8.2511 - acc: 0.475 - ETA: 3s - loss: 8.2343 - acc: 0.476 - ETA: 3s - loss: 8.2134 - acc: 0.477 - ETA: 3s - loss: 8.2304 - acc: 0.476 - ETA: 3s - loss: 8.2472 - acc: 0.476 - ETA: 3s - loss: 8.2745 - acc: 0.474 - ETA: 3s - loss: 8.2780 - acc: 0.474 - ETA: 3s - loss: 8.2867 - acc: 0.473 - ETA: 3s - loss: 8.3000 - acc: 0.472 - ETA: 3s - loss: 8.2985 - acc: 0.472 - ETA: 3s - loss: 8.3045 - acc: 0.472 - ETA: 3s - loss: 8.3003 - acc: 0.472 - ETA: 3s - loss: 8.3022 - acc: 0.472 - ETA: 3s - loss: 8.3137 - acc: 0.471 - ETA: 3s - loss: 8.3063 - acc: 0.472 - ETA: 3s - loss: 8.3110 - acc: 0.471 - ETA: 3s - loss: 8.3099 - acc: 0.472 - ETA: 3s - loss: 8.3123 - acc: 0.471 - ETA: 3s - loss: 8.3146 - acc: 0.471 - ETA: 3s - loss: 8.3136 - acc: 0.471 - ETA: 3s - loss: 8.3093 - acc: 0.472 - ETA: 3s - loss: 8.3149 - acc: 0.472 - ETA: 3s - loss: 8.3042 - acc: 0.472 - ETA: 3s - loss: 8.2968 - acc: 0.473 - ETA: 2s - loss: 8.2958 - acc: 0.473 - ETA: 2s - loss: 8.3016 - acc: 0.472 - ETA: 2s - loss: 8.3071 - acc: 0.472 - ETA: 2s - loss: 8.2965 - acc: 0.473 - ETA: 2s - loss: 8.2959 - acc: 0.473 - ETA: 2s - loss: 8.2892 - acc: 0.473 - ETA: 2s - loss: 8.2843 - acc: 0.473 - ETA: 2s - loss: 8.2818 - acc: 0.473 - ETA: 2s - loss: 8.2866 - acc: 0.472 - ETA: 2s - loss: 8.2677 - acc: 0.473 - ETA: 2s - loss: 8.2575 - acc: 0.474 - ETA: 2s - loss: 8.2563 - acc: 0.474 - ETA: 2s - loss: 8.2559 - acc: 0.474 - ETA: 2s - loss: 8.2398 - acc: 0.475 - ETA: 2s - loss: 8.2409 - acc: 0.475 - ETA: 1s - loss: 8.2448 - acc: 0.475 - ETA: 1s - loss: 8.2521 - acc: 0.475 - ETA: 1s - loss: 8.2767 - acc: 0.473 - ETA: 1s - loss: 8.2805 - acc: 0.473 - ETA: 1s - loss: 8.2672 - acc: 0.474 - ETA: 1s - loss: 8.3019 - acc: 0.472 - ETA: 1s - loss: 8.2968 - acc: 0.472 - ETA: 1s - loss: 8.3111 - acc: 0.471 - ETA: 1s - loss: 8.3189 - acc: 0.471 - ETA: 1s - loss: 8.2978 - acc: 0.472 - ETA: 0s - loss: 8.2936 - acc: 0.473 - ETA: 0s - loss: 8.2983 - acc: 0.472 - ETA: 0s - loss: 8.2941 - acc: 0.473 - ETA: 0s - loss: 8.2862 - acc: 0.473 - ETA: 0s - loss: 8.2798 - acc: 0.473 - ETA: 0s - loss: 8.2812 - acc: 0.473 - ETA: 0s - loss: 8.2773 - acc: 0.473 - ETA: 0s - loss: 8.2835 - acc: 0.473 - ETA: 0s - loss: 8.2753 - acc: 0.473 - ETA: 0s - loss: 8.2816 - acc: 0.473 - ETA: 0s - loss: 8.2813 - acc: 0.473 - ETA: 0s - loss: 8.2780 - acc: 0.473 - ETA: 0s - loss: 8.2800 - acc: 0.473 - ETA: 0s - loss: 8.2818 - acc: 0.473 - ETA: 0s - loss: 8.2836 - acc: 0.473 - ETA: 0s - loss: 8.2758 - acc: 0.473 - ETA: 0s - loss: 8.2729 - acc: 0.473 - ETA: 0s - loss: 8.2821 - acc: 0.473 - ETA: 0s - loss: 8.2821 - acc: 0.4730Epoch 00011: val_loss improved from 8.79615 to 8.72595, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 8.2799 - acc: 0.4731 - val_loss: 8.7259 - val_acc: 0.4084 Epoch 13/20 6660/6680 [============================>.] - ETA: 8s - loss: 10.5074 - acc: 0.35 - ETA: 6s - loss: 9.0753 - acc: 0.4375 - ETA: 6s - loss: 8.9859 - acc: 0.442 - ETA: 6s - loss: 9.0302 - acc: 0.440 - ETA: 6s - loss: 8.4428 - acc: 0.473 - ETA: 6s - loss: 8.3727 - acc: 0.475 - ETA: 6s - loss: 8.3669 - acc: 0.476 - ETA: 6s - loss: 8.6547 - acc: 0.459 - ETA: 5s - loss: 8.5082 - acc: 0.468 - ETA: 5s - loss: 8.3858 - acc: 0.473 - ETA: 5s - loss: 8.3285 - acc: 0.477 - ETA: 5s - loss: 8.3997 - acc: 0.473 - ETA: 5s - loss: 8.2937 - acc: 0.478 - ETA: 6s - loss: 8.2675 - acc: 0.479 - ETA: 6s - loss: 8.3363 - acc: 0.475 - ETA: 6s - loss: 8.3064 - acc: 0.476 - ETA: 6s - loss: 8.3002 - acc: 0.476 - ETA: 6s - loss: 8.2418 - acc: 0.479 - ETA: 6s - loss: 8.1857 - acc: 0.483 - ETA: 6s - loss: 8.1673 - acc: 0.485 - ETA: 6s - loss: 8.0870 - acc: 0.489 - ETA: 7s - loss: 8.1587 - acc: 0.484 - ETA: 7s - loss: 8.1867 - acc: 0.483 - ETA: 7s - loss: 8.1990 - acc: 0.482 - ETA: 7s - loss: 8.1821 - acc: 0.483 - ETA: 8s - loss: 8.1658 - acc: 0.485 - ETA: 8s - loss: 8.1963 - acc: 0.482 - ETA: 8s - loss: 8.1402 - acc: 0.485 - ETA: 9s - loss: 8.1140 - acc: 0.486 - ETA: 9s - loss: 8.0881 - acc: 0.486 - ETA: 9s - loss: 8.1267 - acc: 0.484 - ETA: 9s - loss: 8.1051 - acc: 0.485 - ETA: 9s - loss: 8.2047 - acc: 0.478 - ETA: 9s - loss: 8.1772 - acc: 0.480 - ETA: 9s - loss: 8.2086 - acc: 0.479 - ETA: 8s - loss: 8.1833 - acc: 0.480 - ETA: 8s - loss: 8.1696 - acc: 0.481 - ETA: 8s - loss: 8.1456 - acc: 0.483 - ETA: 8s - loss: 8.1742 - acc: 0.481 - ETA: 8s - loss: 8.1713 - acc: 0.481 - ETA: 8s - loss: 8.2249 - acc: 0.478 - ETA: 8s - loss: 8.2491 - acc: 0.477 - ETA: 7s - loss: 8.2160 - acc: 0.479 - ETA: 7s - loss: 8.2835 - acc: 0.475 - ETA: 7s - loss: 8.2960 - acc: 0.475 - ETA: 7s - loss: 8.2845 - acc: 0.476 - ETA: 7s - loss: 8.2965 - acc: 0.475 - ETA: 7s - loss: 8.2918 - acc: 0.476 - ETA: 7s - loss: 8.2726 - acc: 0.476 - ETA: 7s - loss: 8.2680 - acc: 0.476 - ETA: 7s - loss: 8.2795 - acc: 0.476 - ETA: 7s - loss: 8.2937 - acc: 0.475 - ETA: 7s - loss: 8.2640 - acc: 0.477 - ETA: 6s - loss: 8.2836 - acc: 0.475 - ETA: 6s - loss: 8.2937 - acc: 0.474 - ETA: 6s - loss: 8.3560 - acc: 0.471 - ETA: 6s - loss: 8.3580 - acc: 0.471 - ETA: 6s - loss: 8.3835 - acc: 0.469 - ETA: 6s - loss: 8.3591 - acc: 0.471 - ETA: 6s - loss: 8.3375 - acc: 0.473 - ETA: 6s - loss: 8.3242 - acc: 0.473 - ETA: 6s - loss: 8.3284 - acc: 0.473 - ETA: 6s - loss: 8.3244 - acc: 0.474 - ETA: 6s - loss: 8.3261 - acc: 0.473 - ETA: 6s - loss: 8.3196 - acc: 0.474 - ETA: 6s - loss: 8.3178 - acc: 0.474 - ETA: 6s - loss: 8.3218 - acc: 0.474 - ETA: 6s - loss: 8.3058 - acc: 0.475 - ETA: 6s - loss: 8.2982 - acc: 0.475 - ETA: 6s - loss: 8.3072 - acc: 0.474 - ETA: 6s - loss: 8.3112 - acc: 0.474 - ETA: 6s - loss: 8.3037 - acc: 0.475 - ETA: 6s - loss: 8.3077 - acc: 0.475 - ETA: 6s - loss: 8.3043 - acc: 0.475 - ETA: 6s - loss: 8.2830 - acc: 0.476 - ETA: 6s - loss: 8.3017 - acc: 0.475 - ETA: 6s - loss: 8.2851 - acc: 0.476 - ETA: 5s - loss: 8.2864 - acc: 0.476 - ETA: 5s - loss: 8.2567 - acc: 0.478 - ETA: 5s - loss: 8.2448 - acc: 0.479 - ETA: 5s - loss: 8.2656 - acc: 0.477 - ETA: 5s - loss: 8.2630 - acc: 0.477 - ETA: 5s - loss: 8.2434 - acc: 0.478 - ETA: 5s - loss: 8.2188 - acc: 0.479 - ETA: 5s - loss: 8.2257 - acc: 0.479 - ETA: 5s - loss: 8.2336 - acc: 0.478 - ETA: 5s - loss: 8.2462 - acc: 0.477 - ETA: 5s - loss: 8.2391 - acc: 0.478 - ETA: 4s - loss: 8.2167 - acc: 0.479 - ETA: 4s - loss: 8.1937 - acc: 0.480 - ETA: 4s - loss: 8.1765 - acc: 0.481 - ETA: 4s - loss: 8.1759 - acc: 0.481 - ETA: 4s - loss: 8.1672 - acc: 0.481 - ETA: 4s - loss: 8.1517 - acc: 0.482 - ETA: 4s - loss: 8.1550 - acc: 0.482 - ETA: 4s - loss: 8.1673 - acc: 0.481 - ETA: 4s - loss: 8.1635 - acc: 0.481 - ETA: 4s - loss: 8.1722 - acc: 0.480 - ETA: 4s - loss: 8.1770 - acc: 0.480 - ETA: 4s - loss: 8.1723 - acc: 0.480 - ETA: 4s - loss: 8.1722 - acc: 0.480 - ETA: 4s - loss: 8.1697 - acc: 0.480 - ETA: 4s - loss: 8.1807 - acc: 0.479 - ETA: 4s - loss: 8.2074 - acc: 0.478 - ETA: 4s - loss: 8.2138 - acc: 0.478 - ETA: 4s - loss: 8.2035 - acc: 0.478 - ETA: 4s - loss: 8.1886 - acc: 0.479 - ETA: 4s - loss: 8.1875 - acc: 0.479 - ETA: 3s - loss: 8.1794 - acc: 0.479 - ETA: 3s - loss: 8.1787 - acc: 0.479 - ETA: 3s - loss: 8.1787 - acc: 0.479 - ETA: 3s - loss: 8.1784 - acc: 0.479 - ETA: 3s - loss: 8.1786 - acc: 0.479 - ETA: 3s - loss: 8.1599 - acc: 0.480 - ETA: 3s - loss: 8.1606 - acc: 0.480 - ETA: 3s - loss: 8.1629 - acc: 0.479 - ETA: 3s - loss: 8.1724 - acc: 0.479 - ETA: 3s - loss: 8.1612 - acc: 0.479 - ETA: 3s - loss: 8.1441 - acc: 0.480 - ETA: 3s - loss: 8.1516 - acc: 0.480 - ETA: 3s - loss: 8.1542 - acc: 0.480 - ETA: 2s - loss: 8.1484 - acc: 0.480 - ETA: 2s - loss: 8.1548 - acc: 0.479 - ETA: 2s - loss: 8.1612 - acc: 0.479 - ETA: 2s - loss: 8.1644 - acc: 0.478 - ETA: 2s - loss: 8.1801 - acc: 0.477 - ETA: 2s - loss: 8.1786 - acc: 0.477 - ETA: 2s - loss: 8.1685 - acc: 0.478 - ETA: 2s - loss: 8.1651 - acc: 0.478 - ETA: 2s - loss: 8.1647 - acc: 0.478 - ETA: 2s - loss: 8.1737 - acc: 0.478 - ETA: 2s - loss: 8.1796 - acc: 0.477 - ETA: 2s - loss: 8.1729 - acc: 0.478 - ETA: 2s - loss: 8.1787 - acc: 0.478 - ETA: 2s - loss: 8.1720 - acc: 0.478 - ETA: 2s - loss: 8.1810 - acc: 0.478 - ETA: 2s - loss: 8.1831 - acc: 0.478 - ETA: 2s - loss: 8.1761 - acc: 0.478 - ETA: 2s - loss: 8.1874 - acc: 0.478 - ETA: 2s - loss: 8.1864 - acc: 0.478 - ETA: 2s - loss: 8.1706 - acc: 0.479 - ETA: 2s - loss: 8.1697 - acc: 0.479 - ETA: 2s - loss: 8.1630 - acc: 0.479 - ETA: 1s - loss: 8.1681 - acc: 0.479 - ETA: 1s - loss: 8.1650 - acc: 0.479 - ETA: 1s - loss: 8.1643 - acc: 0.479 - ETA: 1s - loss: 8.1632 - acc: 0.479 - ETA: 1s - loss: 8.1493 - acc: 0.480 - ETA: 1s - loss: 8.1487 - acc: 0.480 - ETA: 1s - loss: 8.1489 - acc: 0.480 - ETA: 1s - loss: 8.1486 - acc: 0.480 - ETA: 1s - loss: 8.1487 - acc: 0.480 - ETA: 1s - loss: 8.1429 - acc: 0.481 - ETA: 1s - loss: 8.1596 - acc: 0.480 - ETA: 1s - loss: 8.1627 - acc: 0.479 - ETA: 0s - loss: 8.1647 - acc: 0.479 - ETA: 0s - loss: 8.1743 - acc: 0.478 - ETA: 0s - loss: 8.1635 - acc: 0.479 - ETA: 0s - loss: 8.1603 - acc: 0.479 - ETA: 0s - loss: 8.1596 - acc: 0.479 - ETA: 0s - loss: 8.1620 - acc: 0.479 - ETA: 0s - loss: 8.1741 - acc: 0.478 - ETA: 0s - loss: 8.1708 - acc: 0.479 - ETA: 0s - loss: 8.1631 - acc: 0.479 - ETA: 0s - loss: 8.1545 - acc: 0.479 - ETA: 0s - loss: 8.1545 - acc: 0.479 - ETA: 0s - loss: 8.1469 - acc: 0.480 - ETA: 0s - loss: 8.1470 - acc: 0.480 - ETA: 0s - loss: 8.1560 - acc: 0.479 - ETA: 0s - loss: 8.1629 - acc: 0.479 - ETA: 0s - loss: 8.1626 - acc: 0.4791Epoch 00012: val_loss improved from 8.72595 to 8.54496, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 8.1623 - acc: 0.4792 - val_loss: 8.5450 - val_acc: 0.4132 Epoch 14/20 6660/6680 [============================>.] - ETA: 7s - loss: 6.4478 - acc: 0.600 - ETA: 6s - loss: 6.5343 - acc: 0.575 - ETA: 6s - loss: 7.3125 - acc: 0.535 - ETA: 6s - loss: 7.4255 - acc: 0.530 - ETA: 6s - loss: 7.2749 - acc: 0.534 - ETA: 6s - loss: 7.6001 - acc: 0.512 - ETA: 6s - loss: 7.6713 - acc: 0.507 - ETA: 6s - loss: 7.8688 - acc: 0.495 - ETA: 6s - loss: 7.8346 - acc: 0.496 - ETA: 6s - loss: 7.9692 - acc: 0.487 - ETA: 6s - loss: 8.0299 - acc: 0.485 - ETA: 6s - loss: 7.9028 - acc: 0.492 - ETA: 6s - loss: 7.8949 - acc: 0.493 - ETA: 6s - loss: 7.8900 - acc: 0.493 - ETA: 6s - loss: 7.8922 - acc: 0.491 - ETA: 6s - loss: 7.9713 - acc: 0.487 - ETA: 6s - loss: 7.9776 - acc: 0.487 - ETA: 6s - loss: 7.9818 - acc: 0.486 - ETA: 6s - loss: 7.9375 - acc: 0.490 - ETA: 6s - loss: 7.9117 - acc: 0.492 - ETA: 6s - loss: 7.9317 - acc: 0.491 - ETA: 6s - loss: 7.9321 - acc: 0.492 - ETA: 6s - loss: 7.9667 - acc: 0.489 - ETA: 6s - loss: 7.9481 - acc: 0.490 - ETA: 6s - loss: 8.0167 - acc: 0.487 - ETA: 6s - loss: 7.9951 - acc: 0.487 - ETA: 6s - loss: 7.9713 - acc: 0.489 - ETA: 6s - loss: 8.0056 - acc: 0.486 - ETA: 6s - loss: 7.9827 - acc: 0.488 - ETA: 7s - loss: 7.9839 - acc: 0.488 - ETA: 7s - loss: 7.9734 - acc: 0.489 - ETA: 7s - loss: 7.9863 - acc: 0.488 - ETA: 7s - loss: 7.9991 - acc: 0.488 - ETA: 7s - loss: 7.9889 - acc: 0.489 - ETA: 7s - loss: 8.0117 - acc: 0.487 - ETA: 7s - loss: 8.0124 - acc: 0.488 - ETA: 8s - loss: 8.0449 - acc: 0.486 - ETA: 8s - loss: 8.0714 - acc: 0.484 - ETA: 8s - loss: 8.1023 - acc: 0.482 - ETA: 8s - loss: 8.0712 - acc: 0.484 - ETA: 8s - loss: 8.0710 - acc: 0.485 - ETA: 8s - loss: 8.0752 - acc: 0.484 - ETA: 8s - loss: 8.0767 - acc: 0.484 - ETA: 7s - loss: 8.0878 - acc: 0.483 - ETA: 7s - loss: 8.1321 - acc: 0.480 - ETA: 7s - loss: 8.1760 - acc: 0.476 - ETA: 7s - loss: 8.1737 - acc: 0.477 - ETA: 7s - loss: 8.1992 - acc: 0.475 - ETA: 7s - loss: 8.1899 - acc: 0.475 - ETA: 7s - loss: 8.1824 - acc: 0.475 - ETA: 7s - loss: 8.1807 - acc: 0.475 - ETA: 7s - loss: 8.1608 - acc: 0.475 - ETA: 7s - loss: 8.0953 - acc: 0.479 - ETA: 7s - loss: 8.0393 - acc: 0.482 - ETA: 7s - loss: 8.0397 - acc: 0.483 - ETA: 6s - loss: 8.0117 - acc: 0.484 - ETA: 6s - loss: 8.0271 - acc: 0.483 - ETA: 6s - loss: 8.0257 - acc: 0.483 - ETA: 6s - loss: 7.9941 - acc: 0.485 - ETA: 6s - loss: 7.9900 - acc: 0.484 - ETA: 6s - loss: 7.9866 - acc: 0.484 - ETA: 6s - loss: 7.9967 - acc: 0.484 - ETA: 6s - loss: 7.9613 - acc: 0.485 - ETA: 6s - loss: 7.9758 - acc: 0.484 - ETA: 6s - loss: 7.9771 - acc: 0.484 - ETA: 6s - loss: 7.9728 - acc: 0.485 - ETA: 5s - loss: 7.9637 - acc: 0.486 - ETA: 5s - loss: 7.9599 - acc: 0.486 - ETA: 5s - loss: 7.9721 - acc: 0.485 - ETA: 5s - loss: 7.9849 - acc: 0.484 - ETA: 5s - loss: 7.9750 - acc: 0.485 - ETA: 5s - loss: 7.9761 - acc: 0.485 - ETA: 5s - loss: 7.9986 - acc: 0.483 - ETA: 5s - loss: 7.9905 - acc: 0.483 - ETA: 5s - loss: 7.9931 - acc: 0.483 - ETA: 5s - loss: 7.9867 - acc: 0.484 - ETA: 5s - loss: 8.0081 - acc: 0.482 - ETA: 5s - loss: 8.0085 - acc: 0.482 - ETA: 5s - loss: 8.0106 - acc: 0.482 - ETA: 5s - loss: 8.0122 - acc: 0.482 - ETA: 5s - loss: 8.0025 - acc: 0.483 - ETA: 5s - loss: 8.0131 - acc: 0.482 - ETA: 5s - loss: 8.0008 - acc: 0.483 - ETA: 5s - loss: 7.9859 - acc: 0.484 - ETA: 5s - loss: 8.0058 - acc: 0.483 - ETA: 5s - loss: 8.0021 - acc: 0.483 - ETA: 5s - loss: 8.0130 - acc: 0.482 - ETA: 5s - loss: 8.0428 - acc: 0.481 - ETA: 5s - loss: 8.0728 - acc: 0.479 - ETA: 5s - loss: 8.0958 - acc: 0.478 - ETA: 5s - loss: 8.0892 - acc: 0.478 - ETA: 5s - loss: 8.0938 - acc: 0.478 - ETA: 5s - loss: 8.0817 - acc: 0.478 - ETA: 4s - loss: 8.0944 - acc: 0.477 - ETA: 4s - loss: 8.0930 - acc: 0.478 - ETA: 4s - loss: 8.1050 - acc: 0.477 - ETA: 4s - loss: 8.1089 - acc: 0.477 - ETA: 4s - loss: 8.0891 - acc: 0.478 - ETA: 4s - loss: 8.0724 - acc: 0.479 - ETA: 4s - loss: 8.0485 - acc: 0.481 - ETA: 4s - loss: 8.0625 - acc: 0.480 - ETA: 4s - loss: 8.0716 - acc: 0.479 - ETA: 4s - loss: 8.0620 - acc: 0.479 - ETA: 4s - loss: 8.0738 - acc: 0.478 - ETA: 3s - loss: 8.0718 - acc: 0.478 - ETA: 3s - loss: 8.0537 - acc: 0.479 - ETA: 3s - loss: 8.0575 - acc: 0.479 - ETA: 3s - loss: 8.0583 - acc: 0.479 - ETA: 3s - loss: 8.0561 - acc: 0.479 - ETA: 3s - loss: 8.0442 - acc: 0.480 - ETA: 3s - loss: 8.0374 - acc: 0.480 - ETA: 3s - loss: 8.0408 - acc: 0.480 - ETA: 3s - loss: 8.0534 - acc: 0.480 - ETA: 3s - loss: 8.0494 - acc: 0.479 - ETA: 3s - loss: 8.0500 - acc: 0.479 - ETA: 3s - loss: 8.0519 - acc: 0.479 - ETA: 3s - loss: 8.0492 - acc: 0.479 - ETA: 3s - loss: 8.0460 - acc: 0.480 - ETA: 3s - loss: 8.0467 - acc: 0.479 - ETA: 3s - loss: 8.0401 - acc: 0.480 - ETA: 3s - loss: 8.0435 - acc: 0.480 - ETA: 3s - loss: 8.0469 - acc: 0.480 - ETA: 3s - loss: 8.0437 - acc: 0.480 - ETA: 3s - loss: 8.0339 - acc: 0.481 - ETA: 3s - loss: 8.0439 - acc: 0.480 - ETA: 2s - loss: 8.0359 - acc: 0.481 - ETA: 2s - loss: 8.0299 - acc: 0.481 - ETA: 2s - loss: 8.0325 - acc: 0.481 - ETA: 2s - loss: 8.0200 - acc: 0.481 - ETA: 2s - loss: 8.0159 - acc: 0.481 - ETA: 2s - loss: 8.0023 - acc: 0.482 - ETA: 2s - loss: 8.0001 - acc: 0.482 - ETA: 2s - loss: 8.0176 - acc: 0.481 - ETA: 2s - loss: 8.0039 - acc: 0.482 - ETA: 2s - loss: 7.9977 - acc: 0.482 - ETA: 2s - loss: 7.9843 - acc: 0.483 - ETA: 2s - loss: 7.9693 - acc: 0.484 - ETA: 2s - loss: 7.9642 - acc: 0.484 - ETA: 2s - loss: 7.9602 - acc: 0.484 - ETA: 2s - loss: 7.9586 - acc: 0.484 - ETA: 2s - loss: 7.9627 - acc: 0.484 - ETA: 1s - loss: 7.9783 - acc: 0.483 - ETA: 1s - loss: 7.9661 - acc: 0.484 - ETA: 1s - loss: 7.9750 - acc: 0.483 - ETA: 1s - loss: 7.9757 - acc: 0.483 - ETA: 1s - loss: 7.9659 - acc: 0.484 - ETA: 1s - loss: 7.9783 - acc: 0.483 - ETA: 1s - loss: 7.9816 - acc: 0.483 - ETA: 1s - loss: 7.9874 - acc: 0.482 - ETA: 1s - loss: 7.9881 - acc: 0.483 - ETA: 1s - loss: 7.9809 - acc: 0.483 - ETA: 1s - loss: 7.9788 - acc: 0.483 - ETA: 0s - loss: 7.9612 - acc: 0.485 - ETA: 0s - loss: 7.9520 - acc: 0.485 - ETA: 0s - loss: 7.9346 - acc: 0.487 - ETA: 0s - loss: 7.9283 - acc: 0.487 - ETA: 0s - loss: 7.9317 - acc: 0.487 - ETA: 0s - loss: 7.9424 - acc: 0.486 - ETA: 0s - loss: 7.9426 - acc: 0.486 - ETA: 0s - loss: 7.9355 - acc: 0.486 - ETA: 0s - loss: 7.9307 - acc: 0.487 - ETA: 0s - loss: 7.9316 - acc: 0.487 - ETA: 0s - loss: 7.9271 - acc: 0.487 - ETA: 0s - loss: 7.9339 - acc: 0.487 - ETA: 0s - loss: 7.9442 - acc: 0.486 - ETA: 0s - loss: 7.9373 - acc: 0.486 - ETA: 0s - loss: 7.9328 - acc: 0.487 - ETA: 0s - loss: 7.9319 - acc: 0.487 - ETA: 0s - loss: 7.9280 - acc: 0.487 - ETA: 0s - loss: 7.9192 - acc: 0.487 - ETA: 0s - loss: 7.9148 - acc: 0.488 - ETA: 0s - loss: 7.9104 - acc: 0.488 - ETA: 0s - loss: 7.9185 - acc: 0.487 - ETA: 0s - loss: 7.9238 - acc: 0.4875Epoch 00013: val_loss improved from 8.54496 to 8.43606, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 7.9226 - acc: 0.4876 - val_loss: 8.4361 - val_acc: 0.4144 Epoch 15/20 6640/6680 [============================>.] - ETA: 8s - loss: 6.4598 - acc: 0.600 - ETA: 6s - loss: 7.0847 - acc: 0.550 - ETA: 6s - loss: 6.8800 - acc: 0.557 - ETA: 6s - loss: 7.6320 - acc: 0.515 - ETA: 6s - loss: 7.9169 - acc: 0.500 - ETA: 6s - loss: 7.8932 - acc: 0.503 - ETA: 6s - loss: 7.6796 - acc: 0.515 - ETA: 6s - loss: 7.4760 - acc: 0.529 - ETA: 6s - loss: 7.5253 - acc: 0.527 - ETA: 6s - loss: 7.3567 - acc: 0.537 - ETA: 6s - loss: 7.3527 - acc: 0.536 - ETA: 6s - loss: 7.3473 - acc: 0.537 - ETA: 6s - loss: 7.2949 - acc: 0.539 - ETA: 6s - loss: 7.3999 - acc: 0.528 - ETA: 7s - loss: 7.4369 - acc: 0.525 - ETA: 7s - loss: 7.4907 - acc: 0.521 - ETA: 7s - loss: 7.5464 - acc: 0.517 - ETA: 7s - loss: 7.5590 - acc: 0.517 - ETA: 8s - loss: 7.6181 - acc: 0.514 - ETA: 8s - loss: 7.6647 - acc: 0.511 - ETA: 8s - loss: 7.5843 - acc: 0.516 - ETA: 9s - loss: 7.5595 - acc: 0.518 - ETA: 9s - loss: 7.5189 - acc: 0.521 - ETA: 9s - loss: 7.5469 - acc: 0.519 - ETA: 10s - loss: 7.5332 - acc: 0.52 - ETA: 10s - loss: 7.5921 - acc: 0.51 - ETA: 10s - loss: 7.6378 - acc: 0.51 - ETA: 10s - loss: 7.6307 - acc: 0.51 - ETA: 11s - loss: 7.6286 - acc: 0.51 - ETA: 10s - loss: 7.6788 - acc: 0.50 - ETA: 10s - loss: 7.6927 - acc: 0.50 - ETA: 10s - loss: 7.7462 - acc: 0.50 - ETA: 10s - loss: 7.6601 - acc: 0.51 - ETA: 9s - loss: 7.6727 - acc: 0.5109 - ETA: 9s - loss: 7.7212 - acc: 0.508 - ETA: 9s - loss: 7.7314 - acc: 0.508 - ETA: 9s - loss: 7.6919 - acc: 0.509 - ETA: 9s - loss: 7.7188 - acc: 0.508 - ETA: 9s - loss: 7.7328 - acc: 0.506 - ETA: 8s - loss: 7.7180 - acc: 0.507 - ETA: 8s - loss: 7.7063 - acc: 0.508 - ETA: 8s - loss: 7.6976 - acc: 0.508 - ETA: 8s - loss: 7.6693 - acc: 0.510 - ETA: 8s - loss: 7.6423 - acc: 0.511 - ETA: 8s - loss: 7.6071 - acc: 0.512 - ETA: 8s - loss: 7.6157 - acc: 0.512 - ETA: 8s - loss: 7.6490 - acc: 0.509 - ETA: 8s - loss: 7.6770 - acc: 0.507 - ETA: 7s - loss: 7.6684 - acc: 0.508 - ETA: 7s - loss: 7.6704 - acc: 0.508 - ETA: 7s - loss: 7.6767 - acc: 0.507 - ETA: 7s - loss: 7.7077 - acc: 0.506 - ETA: 7s - loss: 7.6939 - acc: 0.506 - ETA: 7s - loss: 7.6599 - acc: 0.508 - ETA: 7s - loss: 7.6763 - acc: 0.507 - ETA: 7s - loss: 7.6924 - acc: 0.506 - ETA: 7s - loss: 7.6871 - acc: 0.506 - ETA: 7s - loss: 7.7028 - acc: 0.505 - ETA: 7s - loss: 7.7162 - acc: 0.505 - ETA: 7s - loss: 7.7224 - acc: 0.504 - ETA: 7s - loss: 7.7252 - acc: 0.504 - ETA: 7s - loss: 7.7044 - acc: 0.506 - ETA: 7s - loss: 7.7138 - acc: 0.505 - ETA: 7s - loss: 7.6976 - acc: 0.506 - ETA: 7s - loss: 7.7197 - acc: 0.505 - ETA: 7s - loss: 7.7096 - acc: 0.506 - ETA: 7s - loss: 7.7124 - acc: 0.506 - ETA: 7s - loss: 7.7463 - acc: 0.504 - ETA: 7s - loss: 7.7364 - acc: 0.505 - ETA: 7s - loss: 7.7393 - acc: 0.504 - ETA: 7s - loss: 7.7242 - acc: 0.505 - ETA: 7s - loss: 7.7207 - acc: 0.505 - ETA: 7s - loss: 7.7474 - acc: 0.504 - ETA: 7s - loss: 7.7520 - acc: 0.504 - ETA: 7s - loss: 7.7566 - acc: 0.504 - ETA: 7s - loss: 7.7496 - acc: 0.504 - ETA: 6s - loss: 7.7289 - acc: 0.505 - ETA: 6s - loss: 7.7396 - acc: 0.505 - ETA: 6s - loss: 7.7644 - acc: 0.503 - ETA: 6s - loss: 7.7737 - acc: 0.502 - ETA: 6s - loss: 7.7794 - acc: 0.502 - ETA: 6s - loss: 7.7830 - acc: 0.502 - ETA: 6s - loss: 7.8197 - acc: 0.500 - ETA: 6s - loss: 7.8279 - acc: 0.500 - ETA: 6s - loss: 7.8062 - acc: 0.501 - ETA: 5s - loss: 7.8341 - acc: 0.499 - ETA: 5s - loss: 7.8355 - acc: 0.499 - ETA: 5s - loss: 7.8243 - acc: 0.500 - ETA: 5s - loss: 7.7999 - acc: 0.501 - ETA: 5s - loss: 7.8262 - acc: 0.499 - ETA: 5s - loss: 7.8261 - acc: 0.499 - ETA: 5s - loss: 7.8263 - acc: 0.499 - ETA: 5s - loss: 7.8335 - acc: 0.498 - ETA: 5s - loss: 7.8297 - acc: 0.499 - ETA: 5s - loss: 7.8132 - acc: 0.500 - ETA: 5s - loss: 7.7985 - acc: 0.501 - ETA: 4s - loss: 7.7799 - acc: 0.502 - ETA: 4s - loss: 7.7771 - acc: 0.502 - ETA: 4s - loss: 7.7929 - acc: 0.501 - ETA: 4s - loss: 7.7915 - acc: 0.501 - ETA: 4s - loss: 7.7915 - acc: 0.501 - ETA: 4s - loss: 7.7946 - acc: 0.500 - ETA: 4s - loss: 7.8022 - acc: 0.500 - ETA: 4s - loss: 7.8131 - acc: 0.499 - ETA: 4s - loss: 7.8050 - acc: 0.499 - ETA: 4s - loss: 7.8063 - acc: 0.499 - ETA: 4s - loss: 7.8115 - acc: 0.499 - ETA: 4s - loss: 7.8112 - acc: 0.499 - ETA: 4s - loss: 7.7974 - acc: 0.500 - ETA: 4s - loss: 7.7986 - acc: 0.500 - ETA: 4s - loss: 7.7886 - acc: 0.500 - ETA: 4s - loss: 7.8013 - acc: 0.500 - ETA: 4s - loss: 7.8027 - acc: 0.500 - ETA: 4s - loss: 7.8080 - acc: 0.499 - ETA: 4s - loss: 7.8130 - acc: 0.499 - ETA: 4s - loss: 7.8151 - acc: 0.498 - ETA: 4s - loss: 7.8251 - acc: 0.498 - ETA: 4s - loss: 7.8386 - acc: 0.497 - ETA: 4s - loss: 7.8396 - acc: 0.497 - ETA: 4s - loss: 7.8280 - acc: 0.497 - ETA: 4s - loss: 7.8294 - acc: 0.497 - ETA: 4s - loss: 7.8316 - acc: 0.497 - ETA: 3s - loss: 7.8158 - acc: 0.498 - ETA: 3s - loss: 7.8287 - acc: 0.497 - ETA: 3s - loss: 7.8272 - acc: 0.497 - ETA: 3s - loss: 7.8154 - acc: 0.498 - ETA: 3s - loss: 7.8023 - acc: 0.498 - ETA: 3s - loss: 7.8115 - acc: 0.498 - ETA: 3s - loss: 7.8003 - acc: 0.498 - ETA: 3s - loss: 7.8226 - acc: 0.497 - ETA: 3s - loss: 7.8161 - acc: 0.497 - ETA: 3s - loss: 7.8075 - acc: 0.498 - ETA: 3s - loss: 7.8037 - acc: 0.498 - ETA: 3s - loss: 7.8031 - acc: 0.498 - ETA: 2s - loss: 7.8020 - acc: 0.498 - ETA: 2s - loss: 7.8008 - acc: 0.498 - ETA: 2s - loss: 7.8130 - acc: 0.498 - ETA: 2s - loss: 7.8128 - acc: 0.498 - ETA: 2s - loss: 7.8025 - acc: 0.498 - ETA: 2s - loss: 7.8106 - acc: 0.498 - ETA: 2s - loss: 7.8075 - acc: 0.498 - ETA: 2s - loss: 7.8124 - acc: 0.498 - ETA: 2s - loss: 7.8062 - acc: 0.499 - ETA: 2s - loss: 7.8170 - acc: 0.498 - ETA: 2s - loss: 7.8140 - acc: 0.498 - ETA: 2s - loss: 7.8188 - acc: 0.498 - ETA: 1s - loss: 7.8149 - acc: 0.498 - ETA: 1s - loss: 7.8057 - acc: 0.499 - ETA: 1s - loss: 7.7912 - acc: 0.500 - ETA: 1s - loss: 7.8102 - acc: 0.499 - ETA: 1s - loss: 7.8167 - acc: 0.498 - ETA: 1s - loss: 7.8074 - acc: 0.499 - ETA: 1s - loss: 7.8038 - acc: 0.499 - ETA: 1s - loss: 7.7974 - acc: 0.499 - ETA: 1s - loss: 7.8026 - acc: 0.499 - ETA: 1s - loss: 7.8157 - acc: 0.498 - ETA: 1s - loss: 7.8056 - acc: 0.499 - ETA: 1s - loss: 7.8120 - acc: 0.498 - ETA: 1s - loss: 7.8129 - acc: 0.498 - ETA: 1s - loss: 7.8110 - acc: 0.499 - ETA: 1s - loss: 7.8091 - acc: 0.499 - ETA: 1s - loss: 7.8020 - acc: 0.499 - ETA: 1s - loss: 7.8066 - acc: 0.499 - ETA: 1s - loss: 7.8102 - acc: 0.499 - ETA: 1s - loss: 7.8136 - acc: 0.498 - ETA: 1s - loss: 7.8253 - acc: 0.498 - ETA: 1s - loss: 7.8216 - acc: 0.498 - ETA: 0s - loss: 7.8127 - acc: 0.499 - ETA: 0s - loss: 7.8118 - acc: 0.499 - ETA: 0s - loss: 7.8061 - acc: 0.499 - ETA: 0s - loss: 7.8129 - acc: 0.499 - ETA: 0s - loss: 7.7915 - acc: 0.500 - ETA: 0s - loss: 7.7867 - acc: 0.500 - ETA: 0s - loss: 7.7860 - acc: 0.501 - ETA: 0s - loss: 7.7976 - acc: 0.500 - ETA: 0s - loss: 7.8117 - acc: 0.499 - ETA: 0s - loss: 7.8058 - acc: 0.500 - ETA: 0s - loss: 7.7962 - acc: 0.500 - ETA: 0s - loss: 7.8063 - acc: 0.499 - ETA: 0s - loss: 7.7838 - acc: 0.5009Epoch 00014: val_loss improved from 8.43606 to 8.36570, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 7.7806 - acc: 0.5012 - val_loss: 8.3657 - val_acc: 0.4216 Epoch 16/20 6640/6680 [============================>.] - ETA: 9s - loss: 10.4768 - acc: 0.35 - ETA: 6s - loss: 6.6510 - acc: 0.5875 - ETA: 6s - loss: 7.6513 - acc: 0.514 - ETA: 8s - loss: 7.5009 - acc: 0.525 - ETA: 8s - loss: 7.4224 - acc: 0.525 - ETA: 8s - loss: 7.7443 - acc: 0.504 - ETA: 9s - loss: 7.3874 - acc: 0.528 - ETA: 9s - loss: 7.2172 - acc: 0.540 - ETA: 10s - loss: 7.2720 - acc: 0.53 - ETA: 11s - loss: 7.2813 - acc: 0.53 - ETA: 11s - loss: 7.3474 - acc: 0.53 - ETA: 12s - loss: 7.4273 - acc: 0.52 - ETA: 12s - loss: 7.4589 - acc: 0.52 - ETA: 13s - loss: 7.3340 - acc: 0.53 - ETA: 14s - loss: 7.1850 - acc: 0.54 - ETA: 14s - loss: 7.2580 - acc: 0.53 - ETA: 14s - loss: 7.2577 - acc: 0.54 - ETA: 13s - loss: 7.2275 - acc: 0.54 - ETA: 13s - loss: 7.1496 - acc: 0.54 - ETA: 12s - loss: 7.1825 - acc: 0.54 - ETA: 12s - loss: 7.4268 - acc: 0.53 - ETA: 11s - loss: 7.4565 - acc: 0.52 - ETA: 11s - loss: 7.4018 - acc: 0.53 - ETA: 10s - loss: 7.4371 - acc: 0.53 - ETA: 10s - loss: 7.5478 - acc: 0.52 - ETA: 9s - loss: 7.5313 - acc: 0.5255 - ETA: 9s - loss: 7.5757 - acc: 0.523 - ETA: 9s - loss: 7.5289 - acc: 0.526 - ETA: 8s - loss: 7.6019 - acc: 0.520 - ETA: 8s - loss: 7.6378 - acc: 0.518 - ETA: 8s - loss: 7.6254 - acc: 0.518 - ETA: 8s - loss: 7.6310 - acc: 0.517 - ETA: 8s - loss: 7.6399 - acc: 0.516 - ETA: 7s - loss: 7.6403 - acc: 0.516 - ETA: 7s - loss: 7.6326 - acc: 0.516 - ETA: 7s - loss: 7.6491 - acc: 0.515 - ETA: 7s - loss: 7.6597 - acc: 0.515 - ETA: 7s - loss: 7.6498 - acc: 0.516 - ETA: 7s - loss: 7.6334 - acc: 0.516 - ETA: 7s - loss: 7.6280 - acc: 0.516 - ETA: 7s - loss: 7.6345 - acc: 0.514 - ETA: 7s - loss: 7.6294 - acc: 0.514 - ETA: 7s - loss: 7.5952 - acc: 0.516 - ETA: 7s - loss: 7.6275 - acc: 0.514 - ETA: 7s - loss: 7.6367 - acc: 0.514 - ETA: 7s - loss: 7.6380 - acc: 0.513 - ETA: 7s - loss: 7.6340 - acc: 0.514 - ETA: 7s - loss: 7.6149 - acc: 0.515 - ETA: 6s - loss: 7.6011 - acc: 0.515 - ETA: 6s - loss: 7.5876 - acc: 0.516 - ETA: 6s - loss: 7.5972 - acc: 0.515 - ETA: 6s - loss: 7.5790 - acc: 0.516 - ETA: 6s - loss: 7.5819 - acc: 0.516 - ETA: 6s - loss: 7.5743 - acc: 0.516 - ETA: 7s - loss: 7.5714 - acc: 0.517 - ETA: 7s - loss: 7.5613 - acc: 0.517 - ETA: 7s - loss: 7.5373 - acc: 0.519 - ETA: 7s - loss: 7.5349 - acc: 0.519 - ETA: 7s - loss: 7.5324 - acc: 0.520 - ETA: 7s - loss: 7.5602 - acc: 0.518 - ETA: 7s - loss: 7.5716 - acc: 0.517 - ETA: 7s - loss: 7.5623 - acc: 0.518 - ETA: 7s - loss: 7.5543 - acc: 0.518 - ETA: 7s - loss: 7.5438 - acc: 0.518 - ETA: 7s - loss: 7.5439 - acc: 0.518 - ETA: 7s - loss: 7.5403 - acc: 0.518 - ETA: 7s - loss: 7.5749 - acc: 0.516 - ETA: 7s - loss: 7.5381 - acc: 0.518 - ETA: 7s - loss: 7.5473 - acc: 0.517 - ETA: 7s - loss: 7.5318 - acc: 0.518 - ETA: 7s - loss: 7.5375 - acc: 0.518 - ETA: 6s - loss: 7.5651 - acc: 0.516 - ETA: 6s - loss: 7.5960 - acc: 0.514 - ETA: 6s - loss: 7.6027 - acc: 0.514 - ETA: 6s - loss: 7.6149 - acc: 0.513 - ETA: 6s - loss: 7.5931 - acc: 0.515 - ETA: 6s - loss: 7.6311 - acc: 0.512 - ETA: 6s - loss: 7.6233 - acc: 0.513 - ETA: 6s - loss: 7.6290 - acc: 0.512 - ETA: 6s - loss: 7.5938 - acc: 0.515 - ETA: 6s - loss: 7.5972 - acc: 0.514 - ETA: 5s - loss: 7.5754 - acc: 0.516 - ETA: 5s - loss: 7.5670 - acc: 0.516 - ETA: 5s - loss: 7.5731 - acc: 0.516 - ETA: 5s - loss: 7.5841 - acc: 0.515 - ETA: 5s - loss: 7.5900 - acc: 0.515 - ETA: 5s - loss: 7.5908 - acc: 0.515 - ETA: 5s - loss: 7.5915 - acc: 0.515 - ETA: 5s - loss: 7.6026 - acc: 0.514 - ETA: 5s - loss: 7.5714 - acc: 0.516 - ETA: 5s - loss: 7.5998 - acc: 0.515 - ETA: 5s - loss: 7.5982 - acc: 0.515 - ETA: 5s - loss: 7.6078 - acc: 0.514 - ETA: 4s - loss: 7.6039 - acc: 0.515 - ETA: 4s - loss: 7.6091 - acc: 0.514 - ETA: 4s - loss: 7.5924 - acc: 0.516 - ETA: 4s - loss: 7.5936 - acc: 0.515 - ETA: 4s - loss: 7.6139 - acc: 0.514 - ETA: 4s - loss: 7.6173 - acc: 0.514 - ETA: 4s - loss: 7.6272 - acc: 0.513 - ETA: 4s - loss: 7.6257 - acc: 0.513 - ETA: 4s - loss: 7.6079 - acc: 0.514 - ETA: 4s - loss: 7.6184 - acc: 0.513 - ETA: 4s - loss: 7.6133 - acc: 0.513 - ETA: 4s - loss: 7.6256 - acc: 0.512 - ETA: 4s - loss: 7.6318 - acc: 0.512 - ETA: 4s - loss: 7.6299 - acc: 0.512 - ETA: 4s - loss: 7.6440 - acc: 0.511 - ETA: 4s - loss: 7.6481 - acc: 0.511 - ETA: 4s - loss: 7.6344 - acc: 0.512 - ETA: 4s - loss: 7.6463 - acc: 0.512 - ETA: 4s - loss: 7.6513 - acc: 0.511 - ETA: 4s - loss: 7.6744 - acc: 0.510 - ETA: 4s - loss: 7.6728 - acc: 0.510 - ETA: 4s - loss: 7.6695 - acc: 0.510 - ETA: 4s - loss: 7.6792 - acc: 0.509 - ETA: 3s - loss: 7.6841 - acc: 0.509 - ETA: 3s - loss: 7.6668 - acc: 0.510 - ETA: 3s - loss: 7.6748 - acc: 0.509 - ETA: 3s - loss: 7.6678 - acc: 0.510 - ETA: 3s - loss: 7.6471 - acc: 0.511 - ETA: 3s - loss: 7.6651 - acc: 0.510 - ETA: 3s - loss: 7.6626 - acc: 0.510 - ETA: 3s - loss: 7.6697 - acc: 0.510 - ETA: 3s - loss: 7.6858 - acc: 0.508 - ETA: 3s - loss: 7.6925 - acc: 0.508 - ETA: 3s - loss: 7.6889 - acc: 0.509 - ETA: 3s - loss: 7.7026 - acc: 0.508 - ETA: 3s - loss: 7.7291 - acc: 0.506 - ETA: 2s - loss: 7.7254 - acc: 0.506 - ETA: 2s - loss: 7.7236 - acc: 0.506 - ETA: 2s - loss: 7.7150 - acc: 0.507 - ETA: 2s - loss: 7.7368 - acc: 0.505 - ETA: 2s - loss: 7.7389 - acc: 0.505 - ETA: 2s - loss: 7.7259 - acc: 0.506 - ETA: 2s - loss: 7.7321 - acc: 0.506 - ETA: 2s - loss: 7.7228 - acc: 0.506 - ETA: 2s - loss: 7.7286 - acc: 0.506 - ETA: 2s - loss: 7.7220 - acc: 0.507 - ETA: 2s - loss: 7.7187 - acc: 0.507 - ETA: 2s - loss: 7.7185 - acc: 0.507 - ETA: 2s - loss: 7.7140 - acc: 0.507 - ETA: 2s - loss: 7.7126 - acc: 0.507 - ETA: 1s - loss: 7.7215 - acc: 0.506 - ETA: 1s - loss: 7.7168 - acc: 0.507 - ETA: 1s - loss: 7.7223 - acc: 0.506 - ETA: 1s - loss: 7.7133 - acc: 0.507 - ETA: 1s - loss: 7.6990 - acc: 0.508 - ETA: 1s - loss: 7.6985 - acc: 0.508 - ETA: 1s - loss: 7.6998 - acc: 0.508 - ETA: 1s - loss: 7.7011 - acc: 0.508 - ETA: 1s - loss: 7.6995 - acc: 0.508 - ETA: 1s - loss: 7.7096 - acc: 0.507 - ETA: 1s - loss: 7.7108 - acc: 0.507 - ETA: 1s - loss: 7.7148 - acc: 0.507 - ETA: 1s - loss: 7.7020 - acc: 0.508 - ETA: 1s - loss: 7.6949 - acc: 0.508 - ETA: 1s - loss: 7.6898 - acc: 0.508 - ETA: 1s - loss: 7.7049 - acc: 0.508 - ETA: 1s - loss: 7.7033 - acc: 0.508 - ETA: 1s - loss: 7.7046 - acc: 0.508 - ETA: 1s - loss: 7.7097 - acc: 0.507 - ETA: 1s - loss: 7.7256 - acc: 0.507 - ETA: 1s - loss: 7.7313 - acc: 0.506 - ETA: 1s - loss: 7.7150 - acc: 0.507 - ETA: 1s - loss: 7.7146 - acc: 0.507 - ETA: 0s - loss: 7.7142 - acc: 0.507 - ETA: 0s - loss: 7.7043 - acc: 0.508 - ETA: 0s - loss: 7.7118 - acc: 0.507 - ETA: 0s - loss: 7.7065 - acc: 0.508 - ETA: 0s - loss: 7.7148 - acc: 0.507 - ETA: 0s - loss: 7.7056 - acc: 0.507 - ETA: 0s - loss: 7.7104 - acc: 0.507 - ETA: 0s - loss: 7.7102 - acc: 0.507 - ETA: 0s - loss: 7.7048 - acc: 0.507 - ETA: 0s - loss: 7.7084 - acc: 0.507 - ETA: 0s - loss: 7.7179 - acc: 0.506 - ETA: 0s - loss: 7.7209 - acc: 0.506 - ETA: 0s - loss: 7.7159 - acc: 0.506 - ETA: 0s - loss: 7.7204 - acc: 0.5066Epoch 00015: val_loss improved from 8.36570 to 8.23939, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 7.7167 - acc: 0.5069 - val_loss: 8.2394 - val_acc: 0.4240 Epoch 17/20 6640/6680 [============================>.] - ETA: 25s - loss: 7.2563 - acc: 0.55 - ETA: 22s - loss: 7.2551 - acc: 0.55 - ETA: 23s - loss: 6.9862 - acc: 0.56 - ETA: 21s - loss: 6.6500 - acc: 0.58 - ETA: 18s - loss: 7.5235 - acc: 0.53 - ETA: 15s - loss: 7.3945 - acc: 0.53 - ETA: 14s - loss: 7.1247 - acc: 0.55 - ETA: 13s - loss: 7.0998 - acc: 0.55 - ETA: 12s - loss: 7.3634 - acc: 0.53 - ETA: 11s - loss: 7.1083 - acc: 0.55 - ETA: 10s - loss: 7.4122 - acc: 0.53 - ETA: 10s - loss: 7.3907 - acc: 0.53 - ETA: 9s - loss: 7.5009 - acc: 0.5231 - ETA: 9s - loss: 7.2875 - acc: 0.535 - ETA: 9s - loss: 7.1297 - acc: 0.546 - ETA: 9s - loss: 7.0884 - acc: 0.550 - ETA: 9s - loss: 7.0765 - acc: 0.550 - ETA: 8s - loss: 7.2175 - acc: 0.541 - ETA: 8s - loss: 7.2196 - acc: 0.542 - ETA: 8s - loss: 7.1863 - acc: 0.543 - ETA: 8s - loss: 7.0632 - acc: 0.551 - ETA: 8s - loss: 7.0721 - acc: 0.551 - ETA: 8s - loss: 7.1998 - acc: 0.543 - ETA: 8s - loss: 7.2755 - acc: 0.537 - ETA: 8s - loss: 7.1358 - acc: 0.545 - ETA: 8s - loss: 7.1252 - acc: 0.546 - ETA: 8s - loss: 7.0828 - acc: 0.549 - ETA: 9s - loss: 7.0567 - acc: 0.550 - ETA: 9s - loss: 7.0771 - acc: 0.549 - ETA: 9s - loss: 7.0492 - acc: 0.550 - ETA: 9s - loss: 7.0840 - acc: 0.547 - ETA: 9s - loss: 7.1545 - acc: 0.543 - ETA: 8s - loss: 7.2335 - acc: 0.539 - ETA: 8s - loss: 7.2342 - acc: 0.539 - ETA: 8s - loss: 7.2351 - acc: 0.539 - ETA: 8s - loss: 7.2356 - acc: 0.540 - ETA: 8s - loss: 7.2584 - acc: 0.539 - ETA: 8s - loss: 7.3106 - acc: 0.536 - ETA: 8s - loss: 7.3408 - acc: 0.534 - ETA: 7s - loss: 7.4341 - acc: 0.527 - ETA: 7s - loss: 7.4679 - acc: 0.524 - ETA: 7s - loss: 7.5000 - acc: 0.523 - ETA: 7s - loss: 7.4957 - acc: 0.523 - ETA: 7s - loss: 7.4680 - acc: 0.524 - ETA: 7s - loss: 7.4931 - acc: 0.522 - ETA: 7s - loss: 7.4972 - acc: 0.522 - ETA: 7s - loss: 7.5505 - acc: 0.519 - ETA: 7s - loss: 7.5681 - acc: 0.518 - ETA: 7s - loss: 7.5481 - acc: 0.519 - ETA: 7s - loss: 7.5741 - acc: 0.517 - ETA: 7s - loss: 7.5785 - acc: 0.517 - ETA: 7s - loss: 7.5651 - acc: 0.518 - ETA: 6s - loss: 7.5143 - acc: 0.520 - ETA: 6s - loss: 7.5244 - acc: 0.520 - ETA: 6s - loss: 7.5608 - acc: 0.517 - ETA: 6s - loss: 7.5582 - acc: 0.517 - ETA: 6s - loss: 7.5557 - acc: 0.518 - ETA: 7s - loss: 7.5880 - acc: 0.515 - ETA: 7s - loss: 7.6128 - acc: 0.514 - ETA: 7s - loss: 7.6233 - acc: 0.513 - ETA: 7s - loss: 7.6077 - acc: 0.514 - ETA: 7s - loss: 7.6114 - acc: 0.514 - ETA: 7s - loss: 7.6217 - acc: 0.513 - ETA: 7s - loss: 7.6187 - acc: 0.513 - ETA: 7s - loss: 7.6093 - acc: 0.514 - ETA: 7s - loss: 7.5806 - acc: 0.515 - ETA: 7s - loss: 7.5845 - acc: 0.515 - ETA: 7s - loss: 7.5847 - acc: 0.515 - ETA: 7s - loss: 7.5974 - acc: 0.514 - ETA: 7s - loss: 7.6347 - acc: 0.512 - ETA: 7s - loss: 7.6240 - acc: 0.513 - ETA: 6s - loss: 7.6016 - acc: 0.514 - ETA: 6s - loss: 7.6170 - acc: 0.513 - ETA: 6s - loss: 7.6445 - acc: 0.511 - ETA: 6s - loss: 7.6627 - acc: 0.510 - ETA: 6s - loss: 7.6650 - acc: 0.509 - ETA: 6s - loss: 7.6204 - acc: 0.512 - ETA: 6s - loss: 7.6154 - acc: 0.512 - ETA: 6s - loss: 7.6127 - acc: 0.512 - ETA: 6s - loss: 7.6186 - acc: 0.512 - ETA: 6s - loss: 7.6295 - acc: 0.512 - ETA: 6s - loss: 7.6309 - acc: 0.512 - ETA: 6s - loss: 7.6672 - acc: 0.509 - ETA: 5s - loss: 7.6576 - acc: 0.510 - ETA: 5s - loss: 7.6598 - acc: 0.510 - ETA: 5s - loss: 7.6603 - acc: 0.509 - ETA: 5s - loss: 7.6603 - acc: 0.509 - ETA: 5s - loss: 7.6555 - acc: 0.510 - ETA: 5s - loss: 7.6790 - acc: 0.509 - ETA: 5s - loss: 7.6470 - acc: 0.510 - ETA: 5s - loss: 7.6541 - acc: 0.509 - ETA: 5s - loss: 7.6450 - acc: 0.510 - ETA: 5s - loss: 7.6272 - acc: 0.511 - ETA: 5s - loss: 7.6226 - acc: 0.512 - ETA: 5s - loss: 7.6178 - acc: 0.512 - ETA: 5s - loss: 7.6245 - acc: 0.511 - ETA: 5s - loss: 7.6291 - acc: 0.511 - ETA: 4s - loss: 7.6254 - acc: 0.511 - ETA: 4s - loss: 7.6192 - acc: 0.511 - ETA: 4s - loss: 7.6070 - acc: 0.512 - ETA: 4s - loss: 7.5969 - acc: 0.513 - ETA: 4s - loss: 7.5993 - acc: 0.513 - ETA: 4s - loss: 7.6183 - acc: 0.512 - ETA: 4s - loss: 7.6060 - acc: 0.512 - ETA: 4s - loss: 7.6329 - acc: 0.511 - ETA: 4s - loss: 7.6310 - acc: 0.511 - ETA: 4s - loss: 7.6414 - acc: 0.510 - ETA: 4s - loss: 7.6395 - acc: 0.511 - ETA: 4s - loss: 7.6496 - acc: 0.510 - ETA: 4s - loss: 7.6517 - acc: 0.510 - ETA: 4s - loss: 7.6537 - acc: 0.510 - ETA: 4s - loss: 7.6517 - acc: 0.510 - ETA: 4s - loss: 7.6419 - acc: 0.511 - ETA: 4s - loss: 7.6321 - acc: 0.511 - ETA: 4s - loss: 7.6440 - acc: 0.511 - ETA: 4s - loss: 7.6407 - acc: 0.511 - ETA: 4s - loss: 7.6227 - acc: 0.512 - ETA: 4s - loss: 7.6041 - acc: 0.513 - ETA: 4s - loss: 7.6048 - acc: 0.513 - ETA: 4s - loss: 7.6017 - acc: 0.513 - ETA: 4s - loss: 7.5860 - acc: 0.514 - ETA: 4s - loss: 7.5903 - acc: 0.514 - ETA: 3s - loss: 7.5881 - acc: 0.514 - ETA: 3s - loss: 7.5923 - acc: 0.514 - ETA: 3s - loss: 7.5773 - acc: 0.515 - ETA: 3s - loss: 7.5745 - acc: 0.515 - ETA: 3s - loss: 7.5813 - acc: 0.515 - ETA: 3s - loss: 7.5888 - acc: 0.514 - ETA: 3s - loss: 7.5900 - acc: 0.514 - ETA: 3s - loss: 7.5944 - acc: 0.514 - ETA: 3s - loss: 7.5984 - acc: 0.514 - ETA: 3s - loss: 7.6155 - acc: 0.513 - ETA: 3s - loss: 7.6126 - acc: 0.513 - ETA: 3s - loss: 7.6293 - acc: 0.512 - ETA: 2s - loss: 7.6282 - acc: 0.512 - ETA: 2s - loss: 7.6387 - acc: 0.511 - ETA: 2s - loss: 7.6422 - acc: 0.511 - ETA: 2s - loss: 7.6520 - acc: 0.510 - ETA: 2s - loss: 7.6459 - acc: 0.511 - ETA: 2s - loss: 7.6561 - acc: 0.510 - ETA: 2s - loss: 7.6627 - acc: 0.510 - ETA: 2s - loss: 7.6734 - acc: 0.509 - ETA: 2s - loss: 7.6734 - acc: 0.509 - ETA: 2s - loss: 7.6733 - acc: 0.509 - ETA: 2s - loss: 7.6673 - acc: 0.510 - ETA: 2s - loss: 7.6469 - acc: 0.511 - ETA: 2s - loss: 7.6457 - acc: 0.511 - ETA: 2s - loss: 7.6434 - acc: 0.511 - ETA: 2s - loss: 7.6452 - acc: 0.511 - ETA: 2s - loss: 7.6411 - acc: 0.511 - ETA: 1s - loss: 7.6376 - acc: 0.511 - ETA: 1s - loss: 7.6333 - acc: 0.511 - ETA: 1s - loss: 7.6327 - acc: 0.511 - ETA: 1s - loss: 7.6313 - acc: 0.511 - ETA: 1s - loss: 7.6157 - acc: 0.512 - ETA: 1s - loss: 7.6124 - acc: 0.512 - ETA: 1s - loss: 7.6168 - acc: 0.512 - ETA: 1s - loss: 7.6099 - acc: 0.513 - ETA: 1s - loss: 7.6268 - acc: 0.512 - ETA: 1s - loss: 7.6157 - acc: 0.512 - ETA: 1s - loss: 7.6216 - acc: 0.512 - ETA: 1s - loss: 7.6301 - acc: 0.512 - ETA: 1s - loss: 7.6394 - acc: 0.511 - ETA: 1s - loss: 7.6427 - acc: 0.511 - ETA: 1s - loss: 7.6359 - acc: 0.511 - ETA: 1s - loss: 7.6454 - acc: 0.510 - ETA: 1s - loss: 7.6510 - acc: 0.510 - ETA: 1s - loss: 7.6413 - acc: 0.510 - ETA: 1s - loss: 7.6290 - acc: 0.511 - ETA: 0s - loss: 7.6187 - acc: 0.512 - ETA: 0s - loss: 7.6345 - acc: 0.511 - ETA: 0s - loss: 7.6335 - acc: 0.511 - ETA: 0s - loss: 7.6440 - acc: 0.511 - ETA: 0s - loss: 7.6384 - acc: 0.511 - ETA: 0s - loss: 7.6298 - acc: 0.511 - ETA: 0s - loss: 7.6176 - acc: 0.512 - ETA: 0s - loss: 7.6155 - acc: 0.512 - ETA: 0s - loss: 7.6133 - acc: 0.513 - ETA: 0s - loss: 7.6160 - acc: 0.513 - ETA: 0s - loss: 7.6090 - acc: 0.513 - ETA: 0s - loss: 7.6107 - acc: 0.5134Epoch 00016: val_loss did not improve 6680/6680 [==============================] - 12s - loss: 7.6068 - acc: 0.5136 - val_loss: 8.2560 - val_acc: 0.4251 Epoch 18/20 6640/6680 [============================>.] - ETA: 32s - loss: 5.6414 - acc: 0.65 - ETA: 30s - loss: 5.2387 - acc: 0.67 - ETA: 28s - loss: 5.6417 - acc: 0.65 - ETA: 20s - loss: 5.8250 - acc: 0.63 - ETA: 17s - loss: 6.7041 - acc: 0.57 - ETA: 14s - loss: 7.1736 - acc: 0.54 - ETA: 12s - loss: 7.1387 - acc: 0.55 - ETA: 11s - loss: 7.1612 - acc: 0.55 - ETA: 10s - loss: 7.1128 - acc: 0.55 - ETA: 9s - loss: 7.0288 - acc: 0.5568 - ETA: 9s - loss: 7.1850 - acc: 0.548 - ETA: 8s - loss: 7.3684 - acc: 0.535 - ETA: 8s - loss: 7.3423 - acc: 0.537 - ETA: 8s - loss: 7.4368 - acc: 0.530 - ETA: 8s - loss: 7.4665 - acc: 0.529 - ETA: 8s - loss: 7.5650 - acc: 0.521 - ETA: 7s - loss: 7.5044 - acc: 0.526 - ETA: 7s - loss: 7.3890 - acc: 0.531 - ETA: 7s - loss: 7.4318 - acc: 0.528 - ETA: 7s - loss: 7.4408 - acc: 0.528 - ETA: 7s - loss: 7.4210 - acc: 0.528 - ETA: 7s - loss: 7.4457 - acc: 0.525 - ETA: 7s - loss: 7.4236 - acc: 0.527 - ETA: 6s - loss: 7.4581 - acc: 0.525 - ETA: 6s - loss: 7.5156 - acc: 0.522 - ETA: 6s - loss: 7.5115 - acc: 0.522 - ETA: 6s - loss: 7.4755 - acc: 0.523 - ETA: 6s - loss: 7.5336 - acc: 0.520 - ETA: 6s - loss: 7.5055 - acc: 0.522 - ETA: 6s - loss: 7.5174 - acc: 0.522 - ETA: 6s - loss: 7.5948 - acc: 0.517 - ETA: 6s - loss: 7.6072 - acc: 0.516 - ETA: 6s - loss: 7.5842 - acc: 0.516 - ETA: 6s - loss: 7.5493 - acc: 0.517 - ETA: 6s - loss: 7.5435 - acc: 0.517 - ETA: 6s - loss: 7.5604 - acc: 0.516 - ETA: 6s - loss: 7.5625 - acc: 0.516 - ETA: 6s - loss: 7.5039 - acc: 0.520 - ETA: 6s - loss: 7.5073 - acc: 0.520 - ETA: 6s - loss: 7.5104 - acc: 0.520 - ETA: 6s - loss: 7.5462 - acc: 0.518 - ETA: 6s - loss: 7.5565 - acc: 0.518 - ETA: 6s - loss: 7.5296 - acc: 0.519 - ETA: 6s - loss: 7.5111 - acc: 0.521 - ETA: 6s - loss: 7.4929 - acc: 0.522 - ETA: 6s - loss: 7.4908 - acc: 0.522 - ETA: 6s - loss: 7.4888 - acc: 0.522 - ETA: 6s - loss: 7.5096 - acc: 0.521 - ETA: 6s - loss: 7.5304 - acc: 0.520 - ETA: 6s - loss: 7.5448 - acc: 0.519 - ETA: 6s - loss: 7.5569 - acc: 0.518 - ETA: 7s - loss: 7.5468 - acc: 0.519 - ETA: 7s - loss: 7.5514 - acc: 0.518 - ETA: 7s - loss: 7.5415 - acc: 0.519 - ETA: 7s - loss: 7.5390 - acc: 0.519 - ETA: 7s - loss: 7.5135 - acc: 0.521 - ETA: 7s - loss: 7.5305 - acc: 0.520 - ETA: 6s - loss: 7.5153 - acc: 0.521 - ETA: 6s - loss: 7.5466 - acc: 0.519 - ETA: 6s - loss: 7.5091 - acc: 0.521 - ETA: 6s - loss: 7.5180 - acc: 0.521 - ETA: 6s - loss: 7.5202 - acc: 0.521 - ETA: 6s - loss: 7.5349 - acc: 0.520 - ETA: 6s - loss: 7.5504 - acc: 0.519 - ETA: 6s - loss: 7.5287 - acc: 0.520 - ETA: 6s - loss: 7.5127 - acc: 0.521 - ETA: 6s - loss: 7.5091 - acc: 0.522 - ETA: 6s - loss: 7.5119 - acc: 0.521 - ETA: 6s - loss: 7.5482 - acc: 0.519 - ETA: 5s - loss: 7.5086 - acc: 0.522 - ETA: 5s - loss: 7.5051 - acc: 0.522 - ETA: 5s - loss: 7.4855 - acc: 0.524 - ETA: 5s - loss: 7.5075 - acc: 0.523 - ETA: 5s - loss: 7.5199 - acc: 0.522 - ETA: 5s - loss: 7.5165 - acc: 0.522 - ETA: 5s - loss: 7.5255 - acc: 0.521 - ETA: 5s - loss: 7.5278 - acc: 0.521 - ETA: 5s - loss: 7.5194 - acc: 0.522 - ETA: 5s - loss: 7.4916 - acc: 0.524 - ETA: 5s - loss: 7.4730 - acc: 0.525 - ETA: 5s - loss: 7.4895 - acc: 0.524 - ETA: 5s - loss: 7.5010 - acc: 0.524 - ETA: 4s - loss: 7.5122 - acc: 0.523 - ETA: 4s - loss: 7.5323 - acc: 0.522 - ETA: 4s - loss: 7.5520 - acc: 0.521 - ETA: 4s - loss: 7.5628 - acc: 0.520 - ETA: 4s - loss: 7.5639 - acc: 0.520 - ETA: 4s - loss: 7.5622 - acc: 0.520 - ETA: 4s - loss: 7.5605 - acc: 0.520 - ETA: 4s - loss: 7.5544 - acc: 0.521 - ETA: 4s - loss: 7.5576 - acc: 0.520 - ETA: 4s - loss: 7.5603 - acc: 0.520 - ETA: 4s - loss: 7.5673 - acc: 0.520 - ETA: 4s - loss: 7.5742 - acc: 0.519 - ETA: 4s - loss: 7.5810 - acc: 0.519 - ETA: 4s - loss: 7.5793 - acc: 0.519 - ETA: 4s - loss: 7.5860 - acc: 0.519 - ETA: 4s - loss: 7.5927 - acc: 0.519 - ETA: 4s - loss: 7.5909 - acc: 0.519 - ETA: 4s - loss: 7.5834 - acc: 0.519 - ETA: 4s - loss: 7.5985 - acc: 0.518 - ETA: 4s - loss: 7.5910 - acc: 0.519 - ETA: 4s - loss: 7.6076 - acc: 0.518 - ETA: 4s - loss: 7.5923 - acc: 0.519 - ETA: 4s - loss: 7.5851 - acc: 0.519 - ETA: 4s - loss: 7.5781 - acc: 0.520 - ETA: 4s - loss: 7.5659 - acc: 0.520 - ETA: 4s - loss: 7.5439 - acc: 0.521 - ETA: 3s - loss: 7.5464 - acc: 0.521 - ETA: 3s - loss: 7.5325 - acc: 0.522 - ETA: 3s - loss: 7.5300 - acc: 0.522 - ETA: 3s - loss: 7.5281 - acc: 0.522 - ETA: 3s - loss: 7.5304 - acc: 0.522 - ETA: 3s - loss: 7.5279 - acc: 0.522 - ETA: 3s - loss: 7.5224 - acc: 0.522 - ETA: 3s - loss: 7.4882 - acc: 0.525 - ETA: 3s - loss: 7.4761 - acc: 0.525 - ETA: 3s - loss: 7.4735 - acc: 0.525 - ETA: 3s - loss: 7.4853 - acc: 0.525 - ETA: 3s - loss: 7.4972 - acc: 0.524 - ETA: 3s - loss: 7.5236 - acc: 0.522 - ETA: 2s - loss: 7.5181 - acc: 0.523 - ETA: 2s - loss: 7.5294 - acc: 0.522 - ETA: 2s - loss: 7.5403 - acc: 0.521 - ETA: 2s - loss: 7.5244 - acc: 0.522 - ETA: 2s - loss: 7.5180 - acc: 0.523 - ETA: 2s - loss: 7.5159 - acc: 0.523 - ETA: 2s - loss: 7.5170 - acc: 0.523 - ETA: 2s - loss: 7.5276 - acc: 0.522 - ETA: 2s - loss: 7.5441 - acc: 0.521 - ETA: 2s - loss: 7.5518 - acc: 0.521 - ETA: 2s - loss: 7.5434 - acc: 0.521 - ETA: 2s - loss: 7.5261 - acc: 0.523 - ETA: 2s - loss: 7.5452 - acc: 0.521 - ETA: 2s - loss: 7.5259 - acc: 0.523 - ETA: 2s - loss: 7.5254 - acc: 0.523 - ETA: 2s - loss: 7.5154 - acc: 0.523 - ETA: 1s - loss: 7.5190 - acc: 0.523 - ETA: 1s - loss: 7.5093 - acc: 0.523 - ETA: 1s - loss: 7.5201 - acc: 0.523 - ETA: 1s - loss: 7.5206 - acc: 0.523 - ETA: 1s - loss: 7.5225 - acc: 0.523 - ETA: 1s - loss: 7.5220 - acc: 0.523 - ETA: 1s - loss: 7.5355 - acc: 0.522 - ETA: 1s - loss: 7.5374 - acc: 0.522 - ETA: 1s - loss: 7.5392 - acc: 0.522 - ETA: 1s - loss: 7.5360 - acc: 0.522 - ETA: 1s - loss: 7.5407 - acc: 0.521 - ETA: 1s - loss: 7.5425 - acc: 0.521 - ETA: 1s - loss: 7.5471 - acc: 0.521 - ETA: 1s - loss: 7.5405 - acc: 0.522 - ETA: 1s - loss: 7.5395 - acc: 0.522 - ETA: 1s - loss: 7.5385 - acc: 0.522 - ETA: 1s - loss: 7.5310 - acc: 0.522 - ETA: 1s - loss: 7.5319 - acc: 0.522 - ETA: 1s - loss: 7.5355 - acc: 0.522 - ETA: 1s - loss: 7.5490 - acc: 0.521 - ETA: 1s - loss: 7.5381 - acc: 0.522 - ETA: 1s - loss: 7.5415 - acc: 0.522 - ETA: 1s - loss: 7.5468 - acc: 0.522 - ETA: 0s - loss: 7.5449 - acc: 0.522 - ETA: 0s - loss: 7.5392 - acc: 0.522 - ETA: 0s - loss: 7.5317 - acc: 0.522 - ETA: 0s - loss: 7.5425 - acc: 0.522 - ETA: 0s - loss: 7.5475 - acc: 0.521 - ETA: 0s - loss: 7.5573 - acc: 0.521 - ETA: 0s - loss: 7.5601 - acc: 0.521 - ETA: 0s - loss: 7.5554 - acc: 0.521 - ETA: 0s - loss: 7.5518 - acc: 0.5212Epoch 00017: val_loss improved from 8.23939 to 8.14407, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 12s - loss: 7.5502 - acc: 0.5211 - val_loss: 8.1441 - val_acc: 0.4347 Epoch 19/20 6660/6680 [============================>.] - ETA: 11s - loss: 8.0591 - acc: 0.50 - ETA: 8s - loss: 6.2837 - acc: 0.6000 - ETA: 7s - loss: 6.4058 - acc: 0.585 - ETA: 7s - loss: 6.6627 - acc: 0.575 - ETA: 7s - loss: 7.1091 - acc: 0.550 - ETA: 7s - loss: 7.0252 - acc: 0.553 - ETA: 7s - loss: 7.1226 - acc: 0.544 - ETA: 7s - loss: 7.1431 - acc: 0.545 - ETA: 7s - loss: 7.4593 - acc: 0.527 - ETA: 6s - loss: 7.6525 - acc: 0.514 - ETA: 6s - loss: 7.6457 - acc: 0.515 - ETA: 6s - loss: 7.5624 - acc: 0.519 - ETA: 6s - loss: 7.6038 - acc: 0.518 - ETA: 6s - loss: 7.6810 - acc: 0.511 - ETA: 6s - loss: 7.6697 - acc: 0.513 - ETA: 6s - loss: 7.7012 - acc: 0.511 - ETA: 6s - loss: 7.6736 - acc: 0.513 - ETA: 6s - loss: 7.7378 - acc: 0.510 - ETA: 6s - loss: 7.6756 - acc: 0.512 - ETA: 6s - loss: 7.6899 - acc: 0.512 - ETA: 6s - loss: 7.6601 - acc: 0.514 - ETA: 6s - loss: 7.6884 - acc: 0.512 - ETA: 6s - loss: 7.7221 - acc: 0.510 - ETA: 6s - loss: 7.6475 - acc: 0.513 - ETA: 6s - loss: 7.6225 - acc: 0.515 - ETA: 6s - loss: 7.6490 - acc: 0.514 - ETA: 6s - loss: 7.6614 - acc: 0.514 - ETA: 6s - loss: 7.6198 - acc: 0.516 - ETA: 6s - loss: 7.6503 - acc: 0.515 - ETA: 6s - loss: 7.7022 - acc: 0.512 - ETA: 6s - loss: 7.7073 - acc: 0.512 - ETA: 6s - loss: 7.7018 - acc: 0.511 - ETA: 6s - loss: 7.7186 - acc: 0.510 - ETA: 7s - loss: 7.6905 - acc: 0.512 - ETA: 7s - loss: 7.6666 - acc: 0.513 - ETA: 7s - loss: 7.6505 - acc: 0.514 - ETA: 7s - loss: 7.6663 - acc: 0.513 - ETA: 7s - loss: 7.6914 - acc: 0.512 - ETA: 7s - loss: 7.6670 - acc: 0.513 - ETA: 7s - loss: 7.6360 - acc: 0.514 - ETA: 7s - loss: 7.6092 - acc: 0.516 - ETA: 6s - loss: 7.6600 - acc: 0.513 - ETA: 6s - loss: 7.6949 - acc: 0.512 - ETA: 6s - loss: 7.7066 - acc: 0.511 - ETA: 6s - loss: 7.6764 - acc: 0.513 - ETA: 6s - loss: 7.6718 - acc: 0.514 - ETA: 6s - loss: 7.6753 - acc: 0.514 - ETA: 6s - loss: 7.6293 - acc: 0.517 - ETA: 6s - loss: 7.6190 - acc: 0.518 - ETA: 6s - loss: 7.6390 - acc: 0.517 - ETA: 6s - loss: 7.6623 - acc: 0.515 - ETA: 5s - loss: 7.6449 - acc: 0.516 - ETA: 5s - loss: 7.6231 - acc: 0.517 - ETA: 5s - loss: 7.6473 - acc: 0.516 - ETA: 5s - loss: 7.6520 - acc: 0.515 - ETA: 5s - loss: 7.6620 - acc: 0.515 - ETA: 5s - loss: 7.6561 - acc: 0.515 - ETA: 5s - loss: 7.7012 - acc: 0.512 - ETA: 5s - loss: 7.7124 - acc: 0.512 - ETA: 5s - loss: 7.6807 - acc: 0.513 - ETA: 5s - loss: 7.6461 - acc: 0.516 - ETA: 5s - loss: 7.6519 - acc: 0.515 - ETA: 5s - loss: 7.6444 - acc: 0.516 - ETA: 4s - loss: 7.6639 - acc: 0.515 - ETA: 4s - loss: 7.6316 - acc: 0.517 - ETA: 4s - loss: 7.6052 - acc: 0.518 - ETA: 4s - loss: 7.5849 - acc: 0.520 - ETA: 4s - loss: 7.5755 - acc: 0.521 - ETA: 4s - loss: 7.5663 - acc: 0.521 - ETA: 4s - loss: 7.5735 - acc: 0.521 - ETA: 4s - loss: 7.5811 - acc: 0.520 - ETA: 4s - loss: 7.5576 - acc: 0.522 - ETA: 4s - loss: 7.5092 - acc: 0.524 - ETA: 4s - loss: 7.4925 - acc: 0.525 - ETA: 4s - loss: 7.4814 - acc: 0.526 - ETA: 4s - loss: 7.4992 - acc: 0.525 - ETA: 4s - loss: 7.4930 - acc: 0.525 - ETA: 4s - loss: 7.4869 - acc: 0.526 - ETA: 4s - loss: 7.4862 - acc: 0.526 - ETA: 4s - loss: 7.4965 - acc: 0.525 - ETA: 4s - loss: 7.4951 - acc: 0.525 - ETA: 4s - loss: 7.5122 - acc: 0.524 - ETA: 4s - loss: 7.5107 - acc: 0.524 - ETA: 4s - loss: 7.5092 - acc: 0.524 - ETA: 4s - loss: 7.5123 - acc: 0.524 - ETA: 4s - loss: 7.5277 - acc: 0.523 - ETA: 4s - loss: 7.5336 - acc: 0.523 - ETA: 4s - loss: 7.5272 - acc: 0.523 - ETA: 4s - loss: 7.5546 - acc: 0.522 - ETA: 4s - loss: 7.5561 - acc: 0.521 - ETA: 4s - loss: 7.5699 - acc: 0.521 - ETA: 4s - loss: 7.5777 - acc: 0.520 - ETA: 4s - loss: 7.5607 - acc: 0.521 - ETA: 4s - loss: 7.5440 - acc: 0.522 - ETA: 3s - loss: 7.5411 - acc: 0.523 - ETA: 3s - loss: 7.5415 - acc: 0.523 - ETA: 3s - loss: 7.5494 - acc: 0.522 - ETA: 3s - loss: 7.5491 - acc: 0.522 - ETA: 3s - loss: 7.5534 - acc: 0.522 - ETA: 3s - loss: 7.5506 - acc: 0.522 - ETA: 3s - loss: 7.5776 - acc: 0.520 - ETA: 3s - loss: 7.5698 - acc: 0.521 - ETA: 3s - loss: 7.5597 - acc: 0.521 - ETA: 3s - loss: 7.5632 - acc: 0.521 - ETA: 3s - loss: 7.5607 - acc: 0.521 - ETA: 2s - loss: 7.5643 - acc: 0.521 - ETA: 2s - loss: 7.5706 - acc: 0.521 - ETA: 2s - loss: 7.5583 - acc: 0.522 - ETA: 2s - loss: 7.5511 - acc: 0.522 - ETA: 2s - loss: 7.5520 - acc: 0.522 - ETA: 2s - loss: 7.5529 - acc: 0.522 - ETA: 2s - loss: 7.5362 - acc: 0.523 - ETA: 2s - loss: 7.5458 - acc: 0.523 - ETA: 2s - loss: 7.5468 - acc: 0.523 - ETA: 2s - loss: 7.5482 - acc: 0.522 - ETA: 2s - loss: 7.5527 - acc: 0.522 - ETA: 2s - loss: 7.5441 - acc: 0.523 - ETA: 2s - loss: 7.5565 - acc: 0.522 - ETA: 2s - loss: 7.5666 - acc: 0.521 - ETA: 2s - loss: 7.5796 - acc: 0.520 - ETA: 2s - loss: 7.5875 - acc: 0.520 - ETA: 2s - loss: 7.5893 - acc: 0.520 - ETA: 2s - loss: 7.5790 - acc: 0.521 - ETA: 1s - loss: 7.5808 - acc: 0.521 - ETA: 1s - loss: 7.5803 - acc: 0.520 - ETA: 1s - loss: 7.5889 - acc: 0.520 - ETA: 1s - loss: 7.5946 - acc: 0.519 - ETA: 1s - loss: 7.5964 - acc: 0.519 - ETA: 1s - loss: 7.5893 - acc: 0.520 - ETA: 1s - loss: 7.5721 - acc: 0.521 - ETA: 1s - loss: 7.5657 - acc: 0.521 - ETA: 1s - loss: 7.5663 - acc: 0.521 - ETA: 1s - loss: 7.5670 - acc: 0.521 - ETA: 1s - loss: 7.5705 - acc: 0.521 - ETA: 1s - loss: 7.5869 - acc: 0.520 - ETA: 1s - loss: 7.5795 - acc: 0.520 - ETA: 1s - loss: 7.5923 - acc: 0.520 - ETA: 1s - loss: 7.5999 - acc: 0.519 - ETA: 1s - loss: 7.5896 - acc: 0.520 - ETA: 1s - loss: 7.5897 - acc: 0.520 - ETA: 1s - loss: 7.5740 - acc: 0.521 - ETA: 1s - loss: 7.5774 - acc: 0.520 - ETA: 0s - loss: 7.5715 - acc: 0.521 - ETA: 0s - loss: 7.5658 - acc: 0.521 - ETA: 0s - loss: 7.5545 - acc: 0.522 - ETA: 0s - loss: 7.5672 - acc: 0.521 - ETA: 0s - loss: 7.5598 - acc: 0.522 - ETA: 0s - loss: 7.5482 - acc: 0.522 - ETA: 0s - loss: 7.5515 - acc: 0.522 - ETA: 0s - loss: 7.5572 - acc: 0.522 - ETA: 0s - loss: 7.5395 - acc: 0.523 - ETA: 0s - loss: 7.5280 - acc: 0.524 - ETA: 0s - loss: 7.5124 - acc: 0.525 - ETA: 0s - loss: 7.5158 - acc: 0.525 - ETA: 0s - loss: 7.5229 - acc: 0.524 - ETA: 0s - loss: 7.5214 - acc: 0.5248Epoch 00018: val_loss improved from 8.14407 to 8.11466, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 11s - loss: 7.5255 - acc: 0.5246 - val_loss: 8.1147 - val_acc: 0.4299 Epoch 20/20 6660/6680 [============================>.] - ETA: 21s - loss: 6.4479 - acc: 0.60 - ETA: 14s - loss: 5.7678 - acc: 0.63 - ETA: 10s - loss: 6.0456 - acc: 0.61 - ETA: 9s - loss: 6.7168 - acc: 0.5778 - ETA: 8s - loss: 7.0387 - acc: 0.558 - ETA: 7s - loss: 7.1414 - acc: 0.550 - ETA: 7s - loss: 7.2879 - acc: 0.536 - ETA: 7s - loss: 7.1386 - acc: 0.540 - ETA: 6s - loss: 7.0190 - acc: 0.550 - ETA: 6s - loss: 6.9487 - acc: 0.553 - ETA: 6s - loss: 6.9986 - acc: 0.551 - ETA: 6s - loss: 7.1326 - acc: 0.543 - ETA: 6s - loss: 7.1827 - acc: 0.540 - ETA: 6s - loss: 7.2118 - acc: 0.538 - ETA: 6s - loss: 7.3329 - acc: 0.531 - ETA: 6s - loss: 7.2759 - acc: 0.534 - ETA: 6s - loss: 7.1367 - acc: 0.543 - ETA: 6s - loss: 7.1031 - acc: 0.544 - ETA: 6s - loss: 7.0468 - acc: 0.549 - ETA: 6s - loss: 7.0901 - acc: 0.546 - ETA: 6s - loss: 7.0994 - acc: 0.546 - ETA: 5s - loss: 7.1387 - acc: 0.544 - ETA: 5s - loss: 7.1152 - acc: 0.545 - ETA: 5s - loss: 7.1093 - acc: 0.546 - ETA: 5s - loss: 7.1799 - acc: 0.542 - ETA: 5s - loss: 7.1946 - acc: 0.541 - ETA: 5s - loss: 7.2101 - acc: 0.540 - ETA: 5s - loss: 7.3167 - acc: 0.533 - ETA: 5s - loss: 7.3847 - acc: 0.529 - ETA: 5s - loss: 7.3416 - acc: 0.532 - ETA: 5s - loss: 7.3129 - acc: 0.533 - ETA: 5s - loss: 7.2750 - acc: 0.536 - ETA: 5s - loss: 7.2834 - acc: 0.536 - ETA: 5s - loss: 7.2930 - acc: 0.535 - ETA: 5s - loss: 7.3012 - acc: 0.535 - ETA: 5s - loss: 7.2754 - acc: 0.537 - ETA: 5s - loss: 7.3247 - acc: 0.533 - ETA: 5s - loss: 7.3636 - acc: 0.531 - ETA: 5s - loss: 7.3615 - acc: 0.531 - ETA: 5s - loss: 7.3683 - acc: 0.531 - ETA: 5s - loss: 7.3441 - acc: 0.533 - ETA: 5s - loss: 7.3357 - acc: 0.533 - ETA: 5s - loss: 7.3197 - acc: 0.534 - ETA: 6s - loss: 7.3191 - acc: 0.535 - ETA: 6s - loss: 7.3259 - acc: 0.534 - ETA: 6s - loss: 7.3401 - acc: 0.533 - ETA: 6s - loss: 7.3246 - acc: 0.535 - ETA: 6s - loss: 7.3172 - acc: 0.535 - ETA: 6s - loss: 7.3454 - acc: 0.533 - ETA: 6s - loss: 7.3589 - acc: 0.532 - ETA: 6s - loss: 7.3641 - acc: 0.532 - ETA: 6s - loss: 7.3180 - acc: 0.534 - ETA: 6s - loss: 7.3458 - acc: 0.532 - ETA: 6s - loss: 7.3535 - acc: 0.531 - ETA: 5s - loss: 7.3387 - acc: 0.532 - ETA: 5s - loss: 7.3397 - acc: 0.532 - ETA: 5s - loss: 7.3508 - acc: 0.531 - ETA: 5s - loss: 7.3564 - acc: 0.531 - ETA: 5s - loss: 7.3729 - acc: 0.530 - ETA: 5s - loss: 7.3657 - acc: 0.530 - ETA: 5s - loss: 7.3577 - acc: 0.531 - ETA: 5s - loss: 7.3670 - acc: 0.531 - ETA: 5s - loss: 7.3869 - acc: 0.530 - ETA: 5s - loss: 7.3415 - acc: 0.533 - ETA: 5s - loss: 7.3349 - acc: 0.533 - ETA: 5s - loss: 7.3704 - acc: 0.531 - ETA: 5s - loss: 7.3741 - acc: 0.531 - ETA: 5s - loss: 7.3831 - acc: 0.531 - ETA: 5s - loss: 7.3866 - acc: 0.531 - ETA: 4s - loss: 7.3751 - acc: 0.532 - ETA: 4s - loss: 7.3600 - acc: 0.532 - ETA: 4s - loss: 7.3337 - acc: 0.534 - ETA: 4s - loss: 7.3424 - acc: 0.533 - ETA: 4s - loss: 7.3512 - acc: 0.533 - ETA: 4s - loss: 7.3736 - acc: 0.531 - ETA: 4s - loss: 7.3810 - acc: 0.531 - ETA: 4s - loss: 7.3664 - acc: 0.532 - ETA: 4s - loss: 7.3698 - acc: 0.531 - ETA: 4s - loss: 7.3550 - acc: 0.533 - ETA: 4s - loss: 7.3672 - acc: 0.532 - ETA: 4s - loss: 7.3704 - acc: 0.532 - ETA: 4s - loss: 7.3741 - acc: 0.532 - ETA: 4s - loss: 7.3693 - acc: 0.532 - ETA: 4s - loss: 7.3774 - acc: 0.531 - ETA: 4s - loss: 7.3940 - acc: 0.530 - ETA: 4s - loss: 7.3941 - acc: 0.530 - ETA: 4s - loss: 7.3934 - acc: 0.530 - ETA: 4s - loss: 7.4011 - acc: 0.530 - ETA: 4s - loss: 7.4130 - acc: 0.529 - ETA: 4s - loss: 7.4290 - acc: 0.528 - ETA: 4s - loss: 7.4323 - acc: 0.528 - ETA: 4s - loss: 7.4231 - acc: 0.529 - ETA: 4s - loss: 7.4099 - acc: 0.530 - ETA: 4s - loss: 7.4173 - acc: 0.529 - ETA: 4s - loss: 7.4075 - acc: 0.530 - ETA: 4s - loss: 7.4264 - acc: 0.529 - ETA: 4s - loss: 7.4251 - acc: 0.529 - ETA: 4s - loss: 7.4196 - acc: 0.529 - ETA: 3s - loss: 7.4115 - acc: 0.530 - ETA: 3s - loss: 7.4179 - acc: 0.529 - ETA: 3s - loss: 7.4272 - acc: 0.528 - ETA: 3s - loss: 7.4215 - acc: 0.529 - ETA: 3s - loss: 7.4119 - acc: 0.529 - ETA: 3s - loss: 7.3953 - acc: 0.531 - ETA: 3s - loss: 7.3830 - acc: 0.531 - ETA: 3s - loss: 7.3936 - acc: 0.531 - ETA: 3s - loss: 7.4198 - acc: 0.529 - ETA: 3s - loss: 7.4216 - acc: 0.529 - ETA: 2s - loss: 7.4264 - acc: 0.529 - ETA: 2s - loss: 7.4486 - acc: 0.527 - ETA: 2s - loss: 7.4370 - acc: 0.528 - ETA: 2s - loss: 7.4663 - acc: 0.526 - ETA: 2s - loss: 7.4749 - acc: 0.526 - ETA: 2s - loss: 7.4500 - acc: 0.527 - ETA: 2s - loss: 7.4509 - acc: 0.527 - ETA: 2s - loss: 7.4483 - acc: 0.528 - ETA: 2s - loss: 7.4307 - acc: 0.529 - ETA: 2s - loss: 7.4356 - acc: 0.528 - ETA: 2s - loss: 7.4184 - acc: 0.529 - ETA: 2s - loss: 7.4144 - acc: 0.530 - ETA: 1s - loss: 7.4253 - acc: 0.529 - ETA: 1s - loss: 7.4304 - acc: 0.529 - ETA: 1s - loss: 7.4203 - acc: 0.529 - ETA: 1s - loss: 7.4212 - acc: 0.529 - ETA: 1s - loss: 7.4259 - acc: 0.529 - ETA: 1s - loss: 7.4193 - acc: 0.529 - ETA: 1s - loss: 7.4245 - acc: 0.529 - ETA: 1s - loss: 7.4268 - acc: 0.529 - ETA: 1s - loss: 7.4320 - acc: 0.529 - ETA: 1s - loss: 7.4229 - acc: 0.529 - ETA: 1s - loss: 7.4395 - acc: 0.528 - ETA: 1s - loss: 7.4416 - acc: 0.528 - ETA: 1s - loss: 7.4356 - acc: 0.528 - ETA: 1s - loss: 7.4293 - acc: 0.528 - ETA: 1s - loss: 7.4259 - acc: 0.529 - ETA: 1s - loss: 7.4309 - acc: 0.528 - ETA: 1s - loss: 7.4275 - acc: 0.529 - ETA: 1s - loss: 7.4325 - acc: 0.528 - ETA: 1s - loss: 7.4293 - acc: 0.529 - ETA: 1s - loss: 7.4474 - acc: 0.527 - ETA: 1s - loss: 7.4461 - acc: 0.528 - ETA: 1s - loss: 7.4340 - acc: 0.528 - ETA: 1s - loss: 7.4333 - acc: 0.528 - ETA: 1s - loss: 7.4295 - acc: 0.529 - ETA: 1s - loss: 7.4363 - acc: 0.528 - ETA: 0s - loss: 7.4484 - acc: 0.528 - ETA: 0s - loss: 7.4550 - acc: 0.527 - ETA: 0s - loss: 7.4507 - acc: 0.528 - ETA: 0s - loss: 7.4442 - acc: 0.528 - ETA: 0s - loss: 7.4442 - acc: 0.528 - ETA: 0s - loss: 7.4484 - acc: 0.528 - ETA: 0s - loss: 7.4371 - acc: 0.528 - ETA: 0s - loss: 7.4360 - acc: 0.529 - ETA: 0s - loss: 7.4353 - acc: 0.529 - ETA: 0s - loss: 7.4213 - acc: 0.529 - ETA: 0s - loss: 7.4227 - acc: 0.529 - ETA: 0s - loss: 7.4295 - acc: 0.529 - ETA: 0s - loss: 7.4315 - acc: 0.529 - ETA: 0s - loss: 7.4208 - acc: 0.5299Epoch 00019: val_loss improved from 8.11466 to 8.08270, saving model to saved_models/weights.best.VGG16.hdf5 6680/6680 [==============================] - 11s - loss: 7.4179 - acc: 0.5301 - val_loss: 8.0827 - val_acc: 0.4419
Out[21]:
<keras.callbacks.History at 0x23324867898>
Load the Model with the Best Validation Loss¶
In [22]:
VGG16_model.load_weights('saved_models/weights.best.VGG16.hdf5')
Test the Model¶
Now, we can use the CNN to test how well it identifies breed within our test dataset of dog images. We print the test accuracy below.
In [23]:
# get index of predicted dog breed for each image in test set
VGG16_predictions = [np.argmax(VGG16_model.predict(np.expand_dims(feature, axis=0))) for feature in test_VGG16]
# report test accuracy
test_accuracy = 100*np.sum(np.array(VGG16_predictions)==np.argmax(test_targets, axis=1))/len(VGG16_predictions)
print('Test accuracy: %.4f%%' % test_accuracy)
Test accuracy: 43.6603%
Predict Dog Breed with the Model¶
In [24]:
from extract_bottleneck_features import *
def VGG16_predict_breed(img_path):
# extract bottleneck features
bottleneck_feature = extract_VGG16(path_to_tensor(img_path))
# obtain predicted vector
predicted_vector = VGG16_model.predict(bottleneck_feature)
# return dog breed that is predicted by the model
print('pv:',dog_names[np.argmax(predicted_vector)])
return dog_names[np.argmax(predicted_vector)]
Step 5: Create a CNN to Classify Dog Breeds (using Transfer Learning)¶
You will now use transfer learning to create a CNN that can identify dog breed from images. Your CNN must attain at least 60% accuracy on the test set.
In Step 4, we used transfer learning to create a CNN using VGG-16 bottleneck features. In this section, you must use the bottleneck features from a different pre-trained model. To make things easier for you, we have pre-computed the features for all of the networks that are currently available in Keras:
- VGG-19 bottleneck features
- ResNet-50 bottleneck features
- Inception bottleneck features
- Xception bottleneck features
The files are encoded as such:
Dog{network}Data.npz
where {network}, in the above filename, can be one of VGG19, Resnet50, InceptionV3, or Xception. Pick one of the above architectures, download the corresponding bottleneck features, and store the downloaded file in the bottleneck_features/ folder in the repository.
(IMPLEMENTATION) Obtain Bottleneck Features¶
In the code block below, extract the bottleneck features corresponding to the train, test, and validation sets by running the following:
bottleneck_features = np.load('bottleneck_features/Dog{network}Data.npz')
train_{network} = bottleneck_features['train']
valid_{network} = bottleneck_features['valid']
test_{network} = bottleneck_features['test']In [25]:
### TODO: Obtain bottleneck features from another pre-trained CNN.
bottleneck_features = np.load('bottleneck_features/DogResnet50Data.npz')
train_ResNet50 = bottleneck_features['train']
valid_ResNet50 = bottleneck_features['valid']
test_ResNet50 = bottleneck_features['test']
(IMPLEMENTATION) Model Architecture¶
Create a CNN to classify dog breed. At the end of your code cell block, summarize the layers of your model by executing the line:
<your model's name>.summary()
Question 5: Outline the steps you took to get to your final CNN architecture and your reasoning at each step. Describe why you think the architecture is suitable for the current problem.
Answer:
In [26]:
### TODO: Define your architecture.
Resnet50_model = Sequential()
Resnet50_model.add(GlobalAveragePooling2D(input_shape=train_ResNet50.shape[1:]))
Resnet50_model.add(Dense(133, activation='softmax'))
Resnet50_model.summary()
_________________________________________________________________ Layer (type) Output Shape Param # ================================================================= global_average_pooling2d_2 ( (None, 2048) 0 _________________________________________________________________ dense_4 (Dense) (None, 133) 272517 ================================================================= Total params: 272,517.0 Trainable params: 272,517.0 Non-trainable params: 0.0 _________________________________________________________________
(IMPLEMENTATION) Compile the Model¶
In [27]:
### TODO: Compile the model.
Resnet50_model.compile(loss='categorical_crossentropy', optimizer='rmsprop', metrics=['accuracy'])
(IMPLEMENTATION) Train the Model¶
Train your model in the code cell below. Use model checkpointing to save the model that attains the best validation loss.
You are welcome to augment the training data, but this is not a requirement.
In [28]:
### TODO: Train the model.
checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.ResNet50.hdf5',
verbose=1, save_best_only=True)
Resnet50_model.fit(train_ResNet50, train_targets,
validation_data=(valid_ResNet50, valid_targets),
epochs=20, batch_size=20, callbacks=[checkpointer], verbose=1)
Train on 6680 samples, validate on 835 samples Epoch 1/20 6620/6680 [============================>.] - ETA: 1503s - loss: 5.2674 - acc: 0.0000e+ - ETA: 761s - loss: 6.0713 - acc: 0.0250 - ETA: 382s - loss: 6.2403 - acc: 0.012 - ETA: 256s - loss: 5.9674 - acc: 0.008 - ETA: 193s - loss: 5.8406 - acc: 0.006 - ETA: 156s - loss: 5.6046 - acc: 0.020 - ETA: 143s - loss: 5.5292 - acc: 0.027 - ETA: 133s - loss: 5.5058 - acc: 0.025 - ETA: 124s - loss: 5.4594 - acc: 0.026 - ETA: 108s - loss: 5.3688 - acc: 0.040 - ETA: 95s - loss: 5.2788 - acc: 0.047 - ETA: 86s - loss: 5.1739 - acc: 0.06 - ETA: 78s - loss: 5.1573 - acc: 0.06 - ETA: 71s - loss: 5.0211 - acc: 0.07 - ETA: 63s - loss: 4.9222 - acc: 0.08 - ETA: 59s - loss: 4.8418 - acc: 0.08 - ETA: 55s - loss: 4.7393 - acc: 0.09 - ETA: 52s - loss: 4.6531 - acc: 0.10 - ETA: 49s - loss: 4.5765 - acc: 0.12 - ETA: 48s - loss: 4.5337 - acc: 0.12 - ETA: 46s - loss: 4.4575 - acc: 0.14 - ETA: 45s - loss: 4.4142 - acc: 0.14 - ETA: 41s - loss: 4.3251 - acc: 0.15 - ETA: 41s - loss: 4.2977 - acc: 0.16 - ETA: 40s - loss: 4.2742 - acc: 0.16 - ETA: 40s - loss: 4.2300 - acc: 0.16 - ETA: 39s - loss: 4.2035 - acc: 0.16 - ETA: 38s - loss: 4.1562 - acc: 0.17 - ETA: 38s - loss: 4.1339 - acc: 0.17 - ETA: 36s - loss: 4.0731 - acc: 0.18 - ETA: 35s - loss: 4.0141 - acc: 0.19 - ETA: 35s - loss: 3.9924 - acc: 0.19 - ETA: 34s - loss: 3.9646 - acc: 0.19 - ETA: 34s - loss: 3.9261 - acc: 0.20 - ETA: 33s - loss: 3.8943 - acc: 0.20 - ETA: 32s - loss: 3.8414 - acc: 0.21 - ETA: 31s - loss: 3.7968 - acc: 0.21 - ETA: 29s - loss: 3.7170 - acc: 0.22 - ETA: 28s - loss: 3.6369 - acc: 0.24 - ETA: 26s - loss: 3.5423 - acc: 0.25 - ETA: 25s - loss: 3.4844 - acc: 0.26 - ETA: 24s - loss: 3.4171 - acc: 0.27 - ETA: 22s - loss: 3.3310 - acc: 0.29 - ETA: 21s - loss: 3.2636 - acc: 0.30 - ETA: 20s - loss: 3.1975 - acc: 0.30 - ETA: 19s - loss: 3.1296 - acc: 0.32 - ETA: 18s - loss: 3.0818 - acc: 0.32 - ETA: 17s - loss: 3.0264 - acc: 0.33 - ETA: 17s - loss: 2.9862 - acc: 0.33 - ETA: 16s - loss: 2.9428 - acc: 0.34 - ETA: 15s - loss: 2.9040 - acc: 0.35 - ETA: 15s - loss: 2.8656 - acc: 0.36 - ETA: 14s - loss: 2.8320 - acc: 0.36 - ETA: 14s - loss: 2.7909 - acc: 0.37 - ETA: 13s - loss: 2.7502 - acc: 0.38 - ETA: 13s - loss: 2.7065 - acc: 0.39 - ETA: 12s - loss: 2.6641 - acc: 0.39 - ETA: 12s - loss: 2.6228 - acc: 0.40 - ETA: 11s - loss: 2.5963 - acc: 0.41 - ETA: 11s - loss: 2.5552 - acc: 0.42 - ETA: 10s - loss: 2.5280 - acc: 0.42 - ETA: 10s - loss: 2.4971 - acc: 0.43 - ETA: 10s - loss: 2.4561 - acc: 0.43 - ETA: 9s - loss: 2.4265 - acc: 0.4464 - ETA: 9s - loss: 2.4052 - acc: 0.449 - ETA: 9s - loss: 2.3697 - acc: 0.456 - ETA: 8s - loss: 2.3486 - acc: 0.461 - ETA: 8s - loss: 2.3348 - acc: 0.461 - ETA: 8s - loss: 2.3113 - acc: 0.465 - ETA: 8s - loss: 2.2829 - acc: 0.470 - ETA: 7s - loss: 2.2665 - acc: 0.471 - ETA: 7s - loss: 2.2438 - acc: 0.476 - ETA: 7s - loss: 2.2295 - acc: 0.479 - ETA: 7s - loss: 2.2215 - acc: 0.481 - ETA: 7s - loss: 2.2089 - acc: 0.483 - ETA: 7s - loss: 2.1907 - acc: 0.487 - ETA: 6s - loss: 2.1762 - acc: 0.490 - ETA: 6s - loss: 2.1651 - acc: 0.492 - ETA: 6s - loss: 2.1379 - acc: 0.497 - ETA: 6s - loss: 2.1222 - acc: 0.500 - ETA: 6s - loss: 2.1031 - acc: 0.504 - ETA: 5s - loss: 2.0804 - acc: 0.509 - ETA: 5s - loss: 2.0599 - acc: 0.513 - ETA: 5s - loss: 2.0476 - acc: 0.516 - ETA: 5s - loss: 2.0266 - acc: 0.520 - ETA: 4s - loss: 2.0096 - acc: 0.523 - ETA: 4s - loss: 1.9943 - acc: 0.526 - ETA: 4s - loss: 1.9820 - acc: 0.529 - ETA: 4s - loss: 1.9618 - acc: 0.533 - ETA: 4s - loss: 1.9447 - acc: 0.535 - ETA: 3s - loss: 1.9322 - acc: 0.538 - ETA: 3s - loss: 1.9198 - acc: 0.541 - ETA: 3s - loss: 1.9046 - acc: 0.543 - ETA: 3s - loss: 1.8897 - acc: 0.547 - ETA: 3s - loss: 1.8723 - acc: 0.551 - ETA: 3s - loss: 1.8561 - acc: 0.555 - ETA: 2s - loss: 1.8437 - acc: 0.558 - ETA: 2s - loss: 1.8387 - acc: 0.558 - ETA: 2s - loss: 1.8204 - acc: 0.562 - ETA: 2s - loss: 1.8057 - acc: 0.565 - ETA: 2s - loss: 1.7954 - acc: 0.567 - ETA: 2s - loss: 1.7906 - acc: 0.567 - ETA: 2s - loss: 1.7775 - acc: 0.570 - ETA: 2s - loss: 1.7711 - acc: 0.571 - ETA: 1s - loss: 1.7620 - acc: 0.572 - ETA: 1s - loss: 1.7510 - acc: 0.575 - ETA: 1s - loss: 1.7477 - acc: 0.575 - ETA: 1s - loss: 1.7463 - acc: 0.576 - ETA: 1s - loss: 1.7429 - acc: 0.576 - ETA: 1s - loss: 1.7391 - acc: 0.577 - ETA: 1s - loss: 1.7279 - acc: 0.579 - ETA: 1s - loss: 1.7198 - acc: 0.580 - ETA: 1s - loss: 1.7098 - acc: 0.582 - ETA: 1s - loss: 1.6988 - acc: 0.584 - ETA: 0s - loss: 1.6918 - acc: 0.586 - ETA: 0s - loss: 1.6814 - acc: 0.588 - ETA: 0s - loss: 1.6737 - acc: 0.589 - ETA: 0s - loss: 1.6650 - acc: 0.591 - ETA: 0s - loss: 1.6583 - acc: 0.593 - ETA: 0s - loss: 1.6482 - acc: 0.595 - ETA: 0s - loss: 1.6429 - acc: 0.596 - ETA: 0s - loss: 1.6339 - acc: 0.5982Epoch 00000: val_loss improved from inf to 0.82979, saving model to saved_models/weights.best.ResNet50.hdf5 6680/6680 [==============================] - 12s - loss: 1.6271 - acc: 0.5996 - val_loss: 0.8298 - val_acc: 0.7545 Epoch 2/20 6640/6680 [============================>.] - ETA: 8s - loss: 0.6012 - acc: 0.750 - ETA: 7s - loss: 0.3806 - acc: 0.875 - ETA: 11s - loss: 0.4779 - acc: 0.84 - ETA: 14s - loss: 0.4632 - acc: 0.85 - ETA: 13s - loss: 0.4588 - acc: 0.86 - ETA: 10s - loss: 0.4790 - acc: 0.85 - ETA: 9s - loss: 0.4596 - acc: 0.8625 - ETA: 8s - loss: 0.4909 - acc: 0.847 - ETA: 7s - loss: 0.4919 - acc: 0.845 - ETA: 7s - loss: 0.4744 - acc: 0.846 - ETA: 7s - loss: 0.4620 - acc: 0.848 - ETA: 6s - loss: 0.4443 - acc: 0.854 - ETA: 6s - loss: 0.4405 - acc: 0.859 - ETA: 6s - loss: 0.4419 - acc: 0.862 - ETA: 5s - loss: 0.4272 - acc: 0.868 - ETA: 5s - loss: 0.4170 - acc: 0.872 - ETA: 5s - loss: 0.4287 - acc: 0.871 - ETA: 5s - loss: 0.4282 - acc: 0.869 - ETA: 5s - loss: 0.4290 - acc: 0.869 - ETA: 5s - loss: 0.4418 - acc: 0.865 - ETA: 5s - loss: 0.4436 - acc: 0.864 - ETA: 5s - loss: 0.4354 - acc: 0.865 - ETA: 4s - loss: 0.4373 - acc: 0.862 - ETA: 4s - loss: 0.4553 - acc: 0.860 - ETA: 4s - loss: 0.4497 - acc: 0.862 - ETA: 4s - loss: 0.4445 - acc: 0.862 - ETA: 4s - loss: 0.4417 - acc: 0.863 - ETA: 4s - loss: 0.4442 - acc: 0.861 - ETA: 4s - loss: 0.4414 - acc: 0.864 - ETA: 4s - loss: 0.4401 - acc: 0.865 - ETA: 4s - loss: 0.4369 - acc: 0.866 - ETA: 4s - loss: 0.4322 - acc: 0.869 - ETA: 4s - loss: 0.4317 - acc: 0.869 - ETA: 4s - loss: 0.4309 - acc: 0.869 - ETA: 4s - loss: 0.4323 - acc: 0.868 - ETA: 3s - loss: 0.4296 - acc: 0.868 - ETA: 3s - loss: 0.4307 - acc: 0.867 - ETA: 4s - loss: 0.4319 - acc: 0.866 - ETA: 4s - loss: 0.4312 - acc: 0.866 - ETA: 4s - loss: 0.4306 - acc: 0.866 - ETA: 4s - loss: 0.4309 - acc: 0.865 - ETA: 4s - loss: 0.4317 - acc: 0.866 - ETA: 4s - loss: 0.4289 - acc: 0.866 - ETA: 4s - loss: 0.4278 - acc: 0.866 - ETA: 3s - loss: 0.4286 - acc: 0.866 - ETA: 3s - loss: 0.4281 - acc: 0.867 - ETA: 3s - loss: 0.4318 - acc: 0.866 - ETA: 3s - loss: 0.4361 - acc: 0.865 - ETA: 3s - loss: 0.4360 - acc: 0.865 - ETA: 3s - loss: 0.4329 - acc: 0.866 - ETA: 3s - loss: 0.4294 - acc: 0.867 - ETA: 3s - loss: 0.4331 - acc: 0.866 - ETA: 3s - loss: 0.4337 - acc: 0.866 - ETA: 3s - loss: 0.4387 - acc: 0.864 - ETA: 3s - loss: 0.4358 - acc: 0.864 - ETA: 3s - loss: 0.4349 - acc: 0.865 - ETA: 2s - loss: 0.4348 - acc: 0.865 - ETA: 2s - loss: 0.4366 - acc: 0.864 - ETA: 2s - loss: 0.4369 - acc: 0.864 - ETA: 2s - loss: 0.4363 - acc: 0.864 - ETA: 2s - loss: 0.4336 - acc: 0.865 - ETA: 2s - loss: 0.4356 - acc: 0.864 - ETA: 2s - loss: 0.4359 - acc: 0.864 - ETA: 2s - loss: 0.4374 - acc: 0.864 - ETA: 2s - loss: 0.4404 - acc: 0.863 - ETA: 2s - loss: 0.4407 - acc: 0.863 - ETA: 2s - loss: 0.4409 - acc: 0.863 - ETA: 2s - loss: 0.4403 - acc: 0.863 - ETA: 2s - loss: 0.4380 - acc: 0.864 - ETA: 2s - loss: 0.4370 - acc: 0.864 - ETA: 2s - loss: 0.4356 - acc: 0.865 - ETA: 2s - loss: 0.4386 - acc: 0.864 - ETA: 2s - loss: 0.4374 - acc: 0.864 - ETA: 1s - loss: 0.4360 - acc: 0.865 - ETA: 1s - loss: 0.4389 - acc: 0.864 - ETA: 1s - loss: 0.4396 - acc: 0.863 - ETA: 1s - loss: 0.4410 - acc: 0.863 - ETA: 1s - loss: 0.4405 - acc: 0.862 - ETA: 1s - loss: 0.4395 - acc: 0.862 - ETA: 1s - loss: 0.4396 - acc: 0.861 - ETA: 1s - loss: 0.4392 - acc: 0.861 - ETA: 1s - loss: 0.4376 - acc: 0.862 - ETA: 1s - loss: 0.4383 - acc: 0.862 - ETA: 1s - loss: 0.4368 - acc: 0.863 - ETA: 1s - loss: 0.4390 - acc: 0.862 - ETA: 1s - loss: 0.4387 - acc: 0.862 - ETA: 0s - loss: 0.4371 - acc: 0.863 - ETA: 0s - loss: 0.4363 - acc: 0.863 - ETA: 0s - loss: 0.4375 - acc: 0.862 - ETA: 0s - loss: 0.4380 - acc: 0.862 - ETA: 0s - loss: 0.4400 - acc: 0.862 - ETA: 0s - loss: 0.4409 - acc: 0.862 - ETA: 0s - loss: 0.4408 - acc: 0.862 - ETA: 0s - loss: 0.4413 - acc: 0.862 - ETA: 0s - loss: 0.4424 - acc: 0.862 - ETA: 0s - loss: 0.4426 - acc: 0.861 - ETA: 0s - loss: 0.4465 - acc: 0.861 - ETA: 0s - loss: 0.4444 - acc: 0.861 - ETA: 0s - loss: 0.4444 - acc: 0.861 - ETA: 0s - loss: 0.4433 - acc: 0.861 - ETA: 0s - loss: 0.4433 - acc: 0.8616Epoch 00001: val_loss improved from 0.82979 to 0.65576, saving model to saved_models/weights.best.ResNet50.hdf5 6680/6680 [==============================] - 6s - loss: 0.4446 - acc: 0.8612 - val_loss: 0.6558 - val_acc: 0.7904 Epoch 3/20 6640/6680 [============================>.] - ETA: 5s - loss: 0.4191 - acc: 0.900 - ETA: 11s - loss: 0.3374 - acc: 0.90 - ETA: 13s - loss: 0.2553 - acc: 0.93 - ETA: 8s - loss: 0.2561 - acc: 0.9214 - ETA: 7s - loss: 0.2514 - acc: 0.910 - ETA: 6s - loss: 0.2364 - acc: 0.921 - ETA: 6s - loss: 0.2326 - acc: 0.923 - ETA: 6s - loss: 0.2227 - acc: 0.930 - ETA: 6s - loss: 0.2348 - acc: 0.923 - ETA: 6s - loss: 0.2449 - acc: 0.922 - ETA: 5s - loss: 0.2559 - acc: 0.921 - ETA: 5s - loss: 0.2449 - acc: 0.925 - ETA: 5s - loss: 0.2436 - acc: 0.925 - ETA: 5s - loss: 0.2421 - acc: 0.927 - ETA: 5s - loss: 0.2513 - acc: 0.922 - ETA: 5s - loss: 0.2499 - acc: 0.925 - ETA: 5s - loss: 0.2436 - acc: 0.927 - ETA: 5s - loss: 0.2425 - acc: 0.927 - ETA: 4s - loss: 0.2418 - acc: 0.927 - ETA: 4s - loss: 0.2528 - acc: 0.923 - ETA: 4s - loss: 0.2520 - acc: 0.923 - ETA: 4s - loss: 0.2475 - acc: 0.924 - ETA: 4s - loss: 0.2452 - acc: 0.924 - ETA: 4s - loss: 0.2536 - acc: 0.922 - ETA: 4s - loss: 0.2531 - acc: 0.922 - ETA: 4s - loss: 0.2489 - acc: 0.923 - ETA: 4s - loss: 0.2479 - acc: 0.924 - ETA: 4s - loss: 0.2560 - acc: 0.921 - ETA: 3s - loss: 0.2531 - acc: 0.922 - ETA: 3s - loss: 0.2499 - acc: 0.925 - ETA: 3s - loss: 0.2472 - acc: 0.926 - ETA: 3s - loss: 0.2459 - acc: 0.926 - ETA: 3s - loss: 0.2438 - acc: 0.926 - ETA: 3s - loss: 0.2476 - acc: 0.925 - ETA: 3s - loss: 0.2516 - acc: 0.924 - ETA: 3s - loss: 0.2506 - acc: 0.925 - ETA: 3s - loss: 0.2493 - acc: 0.925 - ETA: 3s - loss: 0.2504 - acc: 0.924 - ETA: 3s - loss: 0.2504 - acc: 0.924 - ETA: 3s - loss: 0.2520 - acc: 0.924 - ETA: 3s - loss: 0.2511 - acc: 0.924 - ETA: 3s - loss: 0.2511 - acc: 0.925 - ETA: 3s - loss: 0.2528 - acc: 0.924 - ETA: 3s - loss: 0.2521 - acc: 0.923 - ETA: 3s - loss: 0.2516 - acc: 0.923 - ETA: 3s - loss: 0.2493 - acc: 0.924 - ETA: 3s - loss: 0.2504 - acc: 0.924 - ETA: 3s - loss: 0.2488 - acc: 0.924 - ETA: 3s - loss: 0.2480 - acc: 0.923 - ETA: 3s - loss: 0.2489 - acc: 0.923 - ETA: 3s - loss: 0.2494 - acc: 0.921 - ETA: 2s - loss: 0.2503 - acc: 0.921 - ETA: 2s - loss: 0.2465 - acc: 0.922 - ETA: 2s - loss: 0.2497 - acc: 0.921 - ETA: 2s - loss: 0.2505 - acc: 0.920 - ETA: 2s - loss: 0.2500 - acc: 0.920 - ETA: 2s - loss: 0.2491 - acc: 0.921 - ETA: 2s - loss: 0.2524 - acc: 0.920 - ETA: 2s - loss: 0.2512 - acc: 0.921 - ETA: 2s - loss: 0.2514 - acc: 0.920 - ETA: 2s - loss: 0.2487 - acc: 0.921 - ETA: 2s - loss: 0.2469 - acc: 0.922 - ETA: 2s - loss: 0.2462 - acc: 0.922 - ETA: 2s - loss: 0.2460 - acc: 0.922 - ETA: 2s - loss: 0.2462 - acc: 0.922 - ETA: 1s - loss: 0.2459 - acc: 0.922 - ETA: 1s - loss: 0.2479 - acc: 0.921 - ETA: 1s - loss: 0.2525 - acc: 0.920 - ETA: 1s - loss: 0.2572 - acc: 0.919 - ETA: 1s - loss: 0.2571 - acc: 0.919 - ETA: 1s - loss: 0.2572 - acc: 0.919 - ETA: 1s - loss: 0.2572 - acc: 0.919 - ETA: 1s - loss: 0.2561 - acc: 0.919 - ETA: 1s - loss: 0.2577 - acc: 0.918 - ETA: 1s - loss: 0.2577 - acc: 0.919 - ETA: 1s - loss: 0.2603 - acc: 0.917 - ETA: 1s - loss: 0.2611 - acc: 0.917 - ETA: 1s - loss: 0.2606 - acc: 0.917 - ETA: 1s - loss: 0.2596 - acc: 0.917 - ETA: 1s - loss: 0.2603 - acc: 0.916 - ETA: 1s - loss: 0.2602 - acc: 0.917 - ETA: 1s - loss: 0.2608 - acc: 0.917 - ETA: 1s - loss: 0.2624 - acc: 0.917 - ETA: 1s - loss: 0.2631 - acc: 0.916 - ETA: 0s - loss: 0.2641 - acc: 0.916 - ETA: 0s - loss: 0.2628 - acc: 0.916 - ETA: 0s - loss: 0.2623 - acc: 0.917 - ETA: 0s - loss: 0.2635 - acc: 0.916 - ETA: 0s - loss: 0.2637 - acc: 0.916 - ETA: 0s - loss: 0.2647 - acc: 0.916 - ETA: 0s - loss: 0.2658 - acc: 0.916 - ETA: 0s - loss: 0.2660 - acc: 0.915 - ETA: 0s - loss: 0.2666 - acc: 0.915 - ETA: 0s - loss: 0.2661 - acc: 0.916 - ETA: 0s - loss: 0.2657 - acc: 0.916 - ETA: 0s - loss: 0.2676 - acc: 0.915 - ETA: 0s - loss: 0.2672 - acc: 0.915 - ETA: 0s - loss: 0.2670 - acc: 0.916 - ETA: 0s - loss: 0.2656 - acc: 0.916 - ETA: 0s - loss: 0.2645 - acc: 0.9169Epoch 00002: val_loss improved from 0.65576 to 0.64222, saving model to saved_models/weights.best.ResNet50.hdf5 6680/6680 [==============================] - 5s - loss: 0.2639 - acc: 0.9168 - val_loss: 0.6422 - val_acc: 0.8144 Epoch 4/20 6620/6680 [============================>.] - ETA: 11s - loss: 0.1509 - acc: 0.95 - ETA: 9s - loss: 0.1611 - acc: 0.9500 - ETA: 10s - loss: 0.1555 - acc: 0.95 - ETA: 10s - loss: 0.1367 - acc: 0.96 - ETA: 8s - loss: 0.1515 - acc: 0.9550 - ETA: 8s - loss: 0.1492 - acc: 0.957 - ETA: 7s - loss: 0.1564 - acc: 0.955 - ETA: 6s - loss: 0.1516 - acc: 0.957 - ETA: 6s - loss: 0.1491 - acc: 0.958 - ETA: 6s - loss: 0.1513 - acc: 0.957 - ETA: 5s - loss: 0.1726 - acc: 0.953 - ETA: 5s - loss: 0.1661 - acc: 0.956 - ETA: 5s - loss: 0.1599 - acc: 0.957 - ETA: 5s - loss: 0.1688 - acc: 0.954 - ETA: 5s - loss: 0.1813 - acc: 0.951 - ETA: 4s - loss: 0.1836 - acc: 0.950 - ETA: 4s - loss: 0.1752 - acc: 0.951 - ETA: 4s - loss: 0.1729 - acc: 0.952 - ETA: 4s - loss: 0.1724 - acc: 0.952 - ETA: 4s - loss: 0.1696 - acc: 0.953 - ETA: 4s - loss: 0.1704 - acc: 0.952 - ETA: 4s - loss: 0.1732 - acc: 0.950 - ETA: 4s - loss: 0.1751 - acc: 0.949 - ETA: 4s - loss: 0.1745 - acc: 0.950 - ETA: 4s - loss: 0.1784 - acc: 0.950 - ETA: 3s - loss: 0.1742 - acc: 0.951 - ETA: 3s - loss: 0.1728 - acc: 0.951 - ETA: 3s - loss: 0.1686 - acc: 0.953 - ETA: 3s - loss: 0.1681 - acc: 0.953 - ETA: 3s - loss: 0.1670 - acc: 0.954 - ETA: 3s - loss: 0.1681 - acc: 0.953 - ETA: 3s - loss: 0.1675 - acc: 0.953 - ETA: 3s - loss: 0.1696 - acc: 0.952 - ETA: 3s - loss: 0.1724 - acc: 0.951 - ETA: 3s - loss: 0.1707 - acc: 0.952 - ETA: 3s - loss: 0.1732 - acc: 0.951 - ETA: 3s - loss: 0.1727 - acc: 0.951 - ETA: 3s - loss: 0.1723 - acc: 0.951 - ETA: 3s - loss: 0.1727 - acc: 0.951 - ETA: 3s - loss: 0.1721 - acc: 0.951 - ETA: 3s - loss: 0.1704 - acc: 0.951 - ETA: 3s - loss: 0.1741 - acc: 0.950 - ETA: 3s - loss: 0.1750 - acc: 0.949 - ETA: 3s - loss: 0.1736 - acc: 0.950 - ETA: 3s - loss: 0.1745 - acc: 0.949 - ETA: 3s - loss: 0.1743 - acc: 0.949 - ETA: 3s - loss: 0.1731 - acc: 0.949 - ETA: 2s - loss: 0.1745 - acc: 0.949 - ETA: 2s - loss: 0.1723 - acc: 0.949 - ETA: 2s - loss: 0.1714 - acc: 0.949 - ETA: 2s - loss: 0.1708 - acc: 0.949 - ETA: 2s - loss: 0.1702 - acc: 0.949 - ETA: 2s - loss: 0.1702 - acc: 0.949 - ETA: 2s - loss: 0.1719 - acc: 0.948 - ETA: 2s - loss: 0.1724 - acc: 0.947 - ETA: 2s - loss: 0.1724 - acc: 0.947 - ETA: 2s - loss: 0.1714 - acc: 0.948 - ETA: 2s - loss: 0.1723 - acc: 0.947 - ETA: 2s - loss: 0.1729 - acc: 0.947 - ETA: 2s - loss: 0.1728 - acc: 0.947 - ETA: 2s - loss: 0.1721 - acc: 0.947 - ETA: 2s - loss: 0.1728 - acc: 0.946 - ETA: 2s - loss: 0.1733 - acc: 0.946 - ETA: 2s - loss: 0.1724 - acc: 0.946 - ETA: 1s - loss: 0.1730 - acc: 0.946 - ETA: 1s - loss: 0.1722 - acc: 0.946 - ETA: 1s - loss: 0.1716 - acc: 0.946 - ETA: 1s - loss: 0.1712 - acc: 0.946 - ETA: 1s - loss: 0.1719 - acc: 0.945 - ETA: 1s - loss: 0.1717 - acc: 0.946 - ETA: 1s - loss: 0.1716 - acc: 0.945 - ETA: 1s - loss: 0.1711 - acc: 0.945 - ETA: 1s - loss: 0.1712 - acc: 0.945 - ETA: 1s - loss: 0.1729 - acc: 0.945 - ETA: 1s - loss: 0.1731 - acc: 0.945 - ETA: 1s - loss: 0.1754 - acc: 0.944 - ETA: 1s - loss: 0.1752 - acc: 0.944 - ETA: 1s - loss: 0.1742 - acc: 0.945 - ETA: 1s - loss: 0.1750 - acc: 0.945 - ETA: 1s - loss: 0.1744 - acc: 0.945 - ETA: 1s - loss: 0.1744 - acc: 0.945 - ETA: 1s - loss: 0.1744 - acc: 0.945 - ETA: 0s - loss: 0.1747 - acc: 0.945 - ETA: 0s - loss: 0.1758 - acc: 0.945 - ETA: 0s - loss: 0.1767 - acc: 0.945 - ETA: 0s - loss: 0.1757 - acc: 0.945 - ETA: 0s - loss: 0.1783 - acc: 0.944 - ETA: 0s - loss: 0.1789 - acc: 0.944 - ETA: 0s - loss: 0.1784 - acc: 0.944 - ETA: 0s - loss: 0.1773 - acc: 0.944 - ETA: 0s - loss: 0.1772 - acc: 0.944 - ETA: 0s - loss: 0.1766 - acc: 0.944 - ETA: 0s - loss: 0.1772 - acc: 0.944 - ETA: 0s - loss: 0.1791 - acc: 0.944 - ETA: 0s - loss: 0.1788 - acc: 0.944 - ETA: 0s - loss: 0.1789 - acc: 0.9437Epoch 00003: val_loss did not improve 6680/6680 [==============================] - 5s - loss: 0.1787 - acc: 0.9439 - val_loss: 0.6448 - val_acc: 0.8144 Epoch 5/20 6660/6680 [============================>.] - ETA: 12s - loss: 0.1505 - acc: 0.95 - ETA: 7s - loss: 0.1243 - acc: 0.9625 - ETA: 7s - loss: 0.0986 - acc: 0.964 - ETA: 6s - loss: 0.0818 - acc: 0.975 - ETA: 6s - loss: 0.0906 - acc: 0.966 - ETA: 7s - loss: 0.0851 - acc: 0.967 - ETA: 7s - loss: 0.0827 - acc: 0.968 - ETA: 7s - loss: 0.0782 - acc: 0.972 - ETA: 7s - loss: 0.0725 - acc: 0.973 - ETA: 6s - loss: 0.0853 - acc: 0.968 - ETA: 6s - loss: 0.0815 - acc: 0.970 - ETA: 6s - loss: 0.0909 - acc: 0.969 - ETA: 5s - loss: 0.0872 - acc: 0.970 - ETA: 5s - loss: 0.0859 - acc: 0.970 - ETA: 5s - loss: 0.0840 - acc: 0.971 - ETA: 5s - loss: 0.0840 - acc: 0.971 - ETA: 5s - loss: 0.0921 - acc: 0.971 - ETA: 4s - loss: 0.0935 - acc: 0.971 - ETA: 4s - loss: 0.0959 - acc: 0.971 - ETA: 4s - loss: 0.1022 - acc: 0.969 - ETA: 4s - loss: 0.1049 - acc: 0.969 - ETA: 4s - loss: 0.1052 - acc: 0.969 - ETA: 4s - loss: 0.1029 - acc: 0.969 - ETA: 4s - loss: 0.1033 - acc: 0.969 - ETA: 4s - loss: 0.1095 - acc: 0.966 - ETA: 4s - loss: 0.1077 - acc: 0.966 - ETA: 3s - loss: 0.1054 - acc: 0.967 - ETA: 3s - loss: 0.1100 - acc: 0.965 - ETA: 3s - loss: 0.1093 - acc: 0.965 - ETA: 3s - loss: 0.1080 - acc: 0.965 - ETA: 3s - loss: 0.1085 - acc: 0.965 - ETA: 3s - loss: 0.1127 - acc: 0.963 - ETA: 3s - loss: 0.1145 - acc: 0.962 - ETA: 3s - loss: 0.1178 - acc: 0.961 - ETA: 3s - loss: 0.1212 - acc: 0.961 - ETA: 3s - loss: 0.1257 - acc: 0.960 - ETA: 3s - loss: 0.1282 - acc: 0.960 - ETA: 3s - loss: 0.1280 - acc: 0.960 - ETA: 3s - loss: 0.1267 - acc: 0.961 - ETA: 3s - loss: 0.1271 - acc: 0.961 - ETA: 3s - loss: 0.1278 - acc: 0.961 - ETA: 3s - loss: 0.1272 - acc: 0.961 - ETA: 3s - loss: 0.1262 - acc: 0.961 - ETA: 3s - loss: 0.1266 - acc: 0.961 - ETA: 3s - loss: 0.1269 - acc: 0.961 - ETA: 2s - loss: 0.1265 - acc: 0.961 - ETA: 2s - loss: 0.1258 - acc: 0.961 - ETA: 2s - loss: 0.1264 - acc: 0.961 - ETA: 2s - loss: 0.1251 - acc: 0.962 - ETA: 2s - loss: 0.1251 - acc: 0.962 - ETA: 2s - loss: 0.1244 - acc: 0.962 - ETA: 2s - loss: 0.1255 - acc: 0.962 - ETA: 2s - loss: 0.1280 - acc: 0.960 - ETA: 2s - loss: 0.1276 - acc: 0.960 - ETA: 2s - loss: 0.1272 - acc: 0.960 - ETA: 2s - loss: 0.1267 - acc: 0.960 - ETA: 2s - loss: 0.1256 - acc: 0.961 - ETA: 2s - loss: 0.1263 - acc: 0.960 - ETA: 2s - loss: 0.1276 - acc: 0.960 - ETA: 2s - loss: 0.1287 - acc: 0.960 - ETA: 1s - loss: 0.1276 - acc: 0.960 - ETA: 1s - loss: 0.1265 - acc: 0.961 - ETA: 1s - loss: 0.1280 - acc: 0.960 - ETA: 1s - loss: 0.1295 - acc: 0.959 - ETA: 1s - loss: 0.1294 - acc: 0.959 - ETA: 1s - loss: 0.1288 - acc: 0.959 - ETA: 1s - loss: 0.1286 - acc: 0.959 - ETA: 1s - loss: 0.1274 - acc: 0.960 - ETA: 1s - loss: 0.1279 - acc: 0.960 - ETA: 1s - loss: 0.1269 - acc: 0.960 - ETA: 1s - loss: 0.1263 - acc: 0.960 - ETA: 1s - loss: 0.1260 - acc: 0.961 - ETA: 1s - loss: 0.1257 - acc: 0.960 - ETA: 1s - loss: 0.1256 - acc: 0.960 - ETA: 1s - loss: 0.1250 - acc: 0.960 - ETA: 1s - loss: 0.1245 - acc: 0.961 - ETA: 1s - loss: 0.1250 - acc: 0.960 - ETA: 1s - loss: 0.1247 - acc: 0.960 - ETA: 1s - loss: 0.1253 - acc: 0.960 - ETA: 1s - loss: 0.1252 - acc: 0.960 - ETA: 1s - loss: 0.1249 - acc: 0.960 - ETA: 1s - loss: 0.1256 - acc: 0.960 - ETA: 1s - loss: 0.1272 - acc: 0.960 - ETA: 1s - loss: 0.1279 - acc: 0.960 - ETA: 1s - loss: 0.1289 - acc: 0.959 - ETA: 1s - loss: 0.1302 - acc: 0.959 - ETA: 0s - loss: 0.1304 - acc: 0.959 - ETA: 0s - loss: 0.1305 - acc: 0.959 - ETA: 0s - loss: 0.1302 - acc: 0.959 - ETA: 0s - loss: 0.1303 - acc: 0.959 - ETA: 0s - loss: 0.1292 - acc: 0.959 - ETA: 0s - loss: 0.1292 - acc: 0.959 - ETA: 0s - loss: 0.1284 - acc: 0.959 - ETA: 0s - loss: 0.1281 - acc: 0.960 - ETA: 0s - loss: 0.1289 - acc: 0.959 - ETA: 0s - loss: 0.1294 - acc: 0.959 - ETA: 0s - loss: 0.1288 - acc: 0.960 - ETA: 0s - loss: 0.1286 - acc: 0.959 - ETA: 0s - loss: 0.1285 - acc: 0.959 - ETA: 0s - loss: 0.1284 - acc: 0.959 - ETA: 0s - loss: 0.1294 - acc: 0.959 - ETA: 0s - loss: 0.1292 - acc: 0.959 - ETA: 0s - loss: 0.1302 - acc: 0.959 - ETA: 0s - loss: 0.1297 - acc: 0.959 - ETA: 0s - loss: 0.1294 - acc: 0.9596Epoch 00004: val_loss did not improve 6680/6680 [==============================] - 6s - loss: 0.1291 - acc: 0.9597 - val_loss: 0.6731 - val_acc: 0.8132 Epoch 6/20 6640/6680 [============================>.] - ETA: 8s - loss: 0.0307 - acc: 1.000 - ETA: 8s - loss: 0.0627 - acc: 0.987 - ETA: 10s - loss: 0.0727 - acc: 0.98 - ETA: 11s - loss: 0.0639 - acc: 0.98 - ETA: 11s - loss: 0.1198 - acc: 0.96 - ETA: 13s - loss: 0.1087 - acc: 0.96 - ETA: 12s - loss: 0.0922 - acc: 0.97 - ETA: 13s - loss: 0.0879 - acc: 0.97 - ETA: 13s - loss: 0.0897 - acc: 0.97 - ETA: 12s - loss: 0.0806 - acc: 0.97 - ETA: 11s - loss: 0.0843 - acc: 0.97 - ETA: 10s - loss: 0.0801 - acc: 0.97 - ETA: 9s - loss: 0.0851 - acc: 0.9740 - ETA: 9s - loss: 0.0935 - acc: 0.973 - ETA: 8s - loss: 0.0926 - acc: 0.972 - ETA: 8s - loss: 0.0953 - acc: 0.972 - ETA: 8s - loss: 0.0954 - acc: 0.973 - ETA: 7s - loss: 0.0915 - acc: 0.973 - ETA: 7s - loss: 0.0915 - acc: 0.974 - ETA: 7s - loss: 0.0911 - acc: 0.973 - ETA: 7s - loss: 0.0885 - acc: 0.974 - ETA: 7s - loss: 0.0885 - acc: 0.975 - ETA: 6s - loss: 0.0894 - acc: 0.973 - ETA: 6s - loss: 0.0902 - acc: 0.972 - ETA: 6s - loss: 0.0886 - acc: 0.972 - ETA: 6s - loss: 0.0885 - acc: 0.972 - ETA: 6s - loss: 0.0880 - acc: 0.973 - ETA: 6s - loss: 0.0871 - acc: 0.973 - ETA: 6s - loss: 0.0856 - acc: 0.974 - ETA: 6s - loss: 0.0855 - acc: 0.974 - ETA: 5s - loss: 0.0853 - acc: 0.973 - ETA: 5s - loss: 0.0877 - acc: 0.973 - ETA: 5s - loss: 0.0856 - acc: 0.974 - ETA: 5s - loss: 0.0871 - acc: 0.974 - ETA: 5s - loss: 0.0855 - acc: 0.974 - ETA: 5s - loss: 0.0852 - acc: 0.974 - ETA: 5s - loss: 0.0855 - acc: 0.974 - ETA: 5s - loss: 0.0866 - acc: 0.973 - ETA: 5s - loss: 0.0858 - acc: 0.974 - ETA: 5s - loss: 0.0859 - acc: 0.973 - ETA: 5s - loss: 0.0857 - acc: 0.973 - ETA: 5s - loss: 0.0867 - acc: 0.973 - ETA: 5s - loss: 0.0866 - acc: 0.973 - ETA: 5s - loss: 0.0864 - acc: 0.973 - ETA: 5s - loss: 0.0859 - acc: 0.973 - ETA: 5s - loss: 0.0854 - acc: 0.973 - ETA: 5s - loss: 0.0842 - acc: 0.973 - ETA: 5s - loss: 0.0831 - acc: 0.974 - ETA: 4s - loss: 0.0839 - acc: 0.974 - ETA: 4s - loss: 0.0855 - acc: 0.973 - ETA: 4s - loss: 0.0843 - acc: 0.973 - ETA: 4s - loss: 0.0836 - acc: 0.973 - ETA: 4s - loss: 0.0846 - acc: 0.973 - ETA: 4s - loss: 0.0878 - acc: 0.972 - ETA: 4s - loss: 0.0887 - acc: 0.971 - ETA: 4s - loss: 0.0903 - acc: 0.971 - ETA: 4s - loss: 0.0886 - acc: 0.972 - ETA: 4s - loss: 0.0890 - acc: 0.972 - ETA: 3s - loss: 0.0892 - acc: 0.972 - ETA: 3s - loss: 0.0880 - acc: 0.972 - ETA: 3s - loss: 0.0875 - acc: 0.972 - ETA: 3s - loss: 0.0873 - acc: 0.973 - ETA: 3s - loss: 0.0883 - acc: 0.972 - ETA: 3s - loss: 0.0886 - acc: 0.972 - ETA: 3s - loss: 0.0893 - acc: 0.972 - ETA: 3s - loss: 0.0898 - acc: 0.971 - ETA: 3s - loss: 0.0894 - acc: 0.971 - ETA: 3s - loss: 0.0890 - acc: 0.972 - ETA: 3s - loss: 0.0886 - acc: 0.972 - ETA: 3s - loss: 0.0880 - acc: 0.972 - ETA: 3s - loss: 0.0876 - acc: 0.972 - ETA: 3s - loss: 0.0879 - acc: 0.972 - ETA: 3s - loss: 0.0877 - acc: 0.972 - ETA: 3s - loss: 0.0882 - acc: 0.972 - ETA: 3s - loss: 0.0879 - acc: 0.972 - ETA: 3s - loss: 0.0880 - acc: 0.972 - ETA: 3s - loss: 0.0883 - acc: 0.972 - ETA: 3s - loss: 0.0883 - acc: 0.972 - ETA: 3s - loss: 0.0894 - acc: 0.972 - ETA: 3s - loss: 0.0890 - acc: 0.972 - ETA: 3s - loss: 0.0883 - acc: 0.972 - ETA: 3s - loss: 0.0877 - acc: 0.973 - ETA: 3s - loss: 0.0876 - acc: 0.973 - ETA: 3s - loss: 0.0870 - acc: 0.973 - ETA: 3s - loss: 0.0869 - acc: 0.973 - ETA: 3s - loss: 0.0865 - acc: 0.973 - ETA: 3s - loss: 0.0861 - acc: 0.973 - ETA: 3s - loss: 0.0867 - acc: 0.972 - ETA: 3s - loss: 0.0870 - acc: 0.972 - ETA: 3s - loss: 0.0886 - acc: 0.972 - ETA: 3s - loss: 0.0884 - acc: 0.972 - ETA: 3s - loss: 0.0883 - acc: 0.972 - ETA: 2s - loss: 0.0878 - acc: 0.972 - ETA: 2s - loss: 0.0872 - acc: 0.973 - ETA: 2s - loss: 0.0870 - acc: 0.972 - ETA: 2s - loss: 0.0863 - acc: 0.973 - ETA: 2s - loss: 0.0863 - acc: 0.972 - ETA: 2s - loss: 0.0857 - acc: 0.973 - ETA: 2s - loss: 0.0865 - acc: 0.972 - ETA: 2s - loss: 0.0867 - acc: 0.972 - ETA: 2s - loss: 0.0862 - acc: 0.972 - ETA: 2s - loss: 0.0860 - acc: 0.972 - ETA: 2s - loss: 0.0857 - acc: 0.972 - ETA: 2s - loss: 0.0854 - acc: 0.972 - ETA: 2s - loss: 0.0853 - acc: 0.972 - ETA: 2s - loss: 0.0861 - acc: 0.972 - ETA: 2s - loss: 0.0867 - acc: 0.972 - ETA: 2s - loss: 0.0864 - acc: 0.972 - ETA: 2s - loss: 0.0857 - acc: 0.972 - ETA: 2s - loss: 0.0855 - acc: 0.972 - ETA: 2s - loss: 0.0859 - acc: 0.972 - ETA: 1s - loss: 0.0854 - acc: 0.972 - ETA: 1s - loss: 0.0853 - acc: 0.972 - ETA: 1s - loss: 0.0857 - acc: 0.971 - ETA: 1s - loss: 0.0853 - acc: 0.971 - ETA: 1s - loss: 0.0855 - acc: 0.971 - ETA: 1s - loss: 0.0851 - acc: 0.971 - ETA: 1s - loss: 0.0849 - acc: 0.971 - ETA: 1s - loss: 0.0851 - acc: 0.971 - ETA: 1s - loss: 0.0853 - acc: 0.971 - ETA: 1s - loss: 0.0856 - acc: 0.971 - ETA: 1s - loss: 0.0848 - acc: 0.972 - ETA: 1s - loss: 0.0845 - acc: 0.972 - ETA: 0s - loss: 0.0848 - acc: 0.972 - ETA: 0s - loss: 0.0854 - acc: 0.971 - ETA: 0s - loss: 0.0856 - acc: 0.971 - ETA: 0s - loss: 0.0855 - acc: 0.971 - ETA: 0s - loss: 0.0857 - acc: 0.971 - ETA: 0s - loss: 0.0856 - acc: 0.971 - ETA: 0s - loss: 0.0863 - acc: 0.971 - ETA: 0s - loss: 0.0860 - acc: 0.971 - ETA: 0s - loss: 0.0864 - acc: 0.971 - ETA: 0s - loss: 0.0862 - acc: 0.971 - ETA: 0s - loss: 0.0856 - acc: 0.971 - ETA: 0s - loss: 0.0856 - acc: 0.971 - ETA: 0s - loss: 0.0856 - acc: 0.971 - ETA: 0s - loss: 0.0859 - acc: 0.971 - ETA: 0s - loss: 0.0857 - acc: 0.971 - ETA: 0s - loss: 0.0855 - acc: 0.971 - ETA: 0s - loss: 0.0858 - acc: 0.971 - ETA: 0s - loss: 0.0858 - acc: 0.9715Epoch 00005: val_loss did not improve 6680/6680 [==============================] - 8s - loss: 0.0862 - acc: 0.9714 - val_loss: 0.6833 - val_acc: 0.8204 Epoch 7/20 6660/6680 [============================>.] - ETA: 6s - loss: 0.0225 - acc: 1.000 - ETA: 5s - loss: 0.1017 - acc: 0.987 - ETA: 5s - loss: 0.0694 - acc: 0.987 - ETA: 5s - loss: 0.0809 - acc: 0.981 - ETA: 5s - loss: 0.0786 - acc: 0.978 - ETA: 5s - loss: 0.0781 - acc: 0.976 - ETA: 5s - loss: 0.0695 - acc: 0.980 - ETA: 5s - loss: 0.0676 - acc: 0.982 - ETA: 5s - loss: 0.0686 - acc: 0.982 - ETA: 5s - loss: 0.0654 - acc: 0.982 - ETA: 5s - loss: 0.0601 - acc: 0.984 - ETA: 5s - loss: 0.0586 - acc: 0.984 - ETA: 5s - loss: 0.0556 - acc: 0.985 - ETA: 5s - loss: 0.0544 - acc: 0.985 - ETA: 5s - loss: 0.0532 - acc: 0.985 - ETA: 4s - loss: 0.0549 - acc: 0.985 - ETA: 4s - loss: 0.0563 - acc: 0.985 - ETA: 4s - loss: 0.0547 - acc: 0.986 - ETA: 4s - loss: 0.0529 - acc: 0.987 - ETA: 5s - loss: 0.0520 - acc: 0.987 - ETA: 4s - loss: 0.0512 - acc: 0.987 - ETA: 5s - loss: 0.0506 - acc: 0.987 - ETA: 5s - loss: 0.0497 - acc: 0.988 - ETA: 5s - loss: 0.0501 - acc: 0.987 - ETA: 5s - loss: 0.0489 - acc: 0.988 - ETA: 5s - loss: 0.0509 - acc: 0.987 - ETA: 5s - loss: 0.0515 - acc: 0.987 - ETA: 5s - loss: 0.0511 - acc: 0.987 - ETA: 5s - loss: 0.0512 - acc: 0.987 - ETA: 5s - loss: 0.0520 - acc: 0.986 - ETA: 5s - loss: 0.0522 - acc: 0.986 - ETA: 5s - loss: 0.0529 - acc: 0.986 - ETA: 5s - loss: 0.0525 - acc: 0.987 - ETA: 5s - loss: 0.0527 - acc: 0.987 - ETA: 5s - loss: 0.0532 - acc: 0.986 - ETA: 4s - loss: 0.0521 - acc: 0.987 - ETA: 4s - loss: 0.0513 - acc: 0.987 - ETA: 4s - loss: 0.0505 - acc: 0.988 - ETA: 4s - loss: 0.0500 - acc: 0.988 - ETA: 4s - loss: 0.0519 - acc: 0.987 - ETA: 4s - loss: 0.0523 - acc: 0.987 - ETA: 4s - loss: 0.0528 - acc: 0.987 - ETA: 4s - loss: 0.0523 - acc: 0.987 - ETA: 4s - loss: 0.0518 - acc: 0.987 - ETA: 4s - loss: 0.0528 - acc: 0.986 - ETA: 4s - loss: 0.0539 - acc: 0.986 - ETA: 4s - loss: 0.0551 - acc: 0.986 - ETA: 3s - loss: 0.0547 - acc: 0.986 - ETA: 3s - loss: 0.0543 - acc: 0.986 - ETA: 3s - loss: 0.0551 - acc: 0.986 - ETA: 3s - loss: 0.0550 - acc: 0.986 - ETA: 3s - loss: 0.0546 - acc: 0.986 - ETA: 3s - loss: 0.0563 - acc: 0.986 - ETA: 3s - loss: 0.0585 - acc: 0.986 - ETA: 3s - loss: 0.0604 - acc: 0.985 - ETA: 3s - loss: 0.0600 - acc: 0.986 - ETA: 3s - loss: 0.0606 - acc: 0.985 - ETA: 3s - loss: 0.0619 - acc: 0.984 - ETA: 3s - loss: 0.0620 - acc: 0.984 - ETA: 3s - loss: 0.0618 - acc: 0.984 - ETA: 3s - loss: 0.0616 - acc: 0.984 - ETA: 3s - loss: 0.0619 - acc: 0.983 - ETA: 3s - loss: 0.0614 - acc: 0.983 - ETA: 3s - loss: 0.0617 - acc: 0.983 - ETA: 3s - loss: 0.0610 - acc: 0.983 - ETA: 3s - loss: 0.0607 - acc: 0.983 - ETA: 3s - loss: 0.0606 - acc: 0.983 - ETA: 3s - loss: 0.0606 - acc: 0.983 - ETA: 2s - loss: 0.0619 - acc: 0.983 - ETA: 2s - loss: 0.0647 - acc: 0.982 - ETA: 2s - loss: 0.0641 - acc: 0.982 - ETA: 2s - loss: 0.0645 - acc: 0.982 - ETA: 2s - loss: 0.0645 - acc: 0.982 - ETA: 2s - loss: 0.0645 - acc: 0.982 - ETA: 2s - loss: 0.0642 - acc: 0.982 - ETA: 2s - loss: 0.0635 - acc: 0.982 - ETA: 2s - loss: 0.0632 - acc: 0.982 - ETA: 2s - loss: 0.0635 - acc: 0.982 - ETA: 2s - loss: 0.0628 - acc: 0.982 - ETA: 2s - loss: 0.0622 - acc: 0.982 - ETA: 2s - loss: 0.0621 - acc: 0.982 - ETA: 2s - loss: 0.0654 - acc: 0.982 - ETA: 1s - loss: 0.0660 - acc: 0.982 - ETA: 1s - loss: 0.0661 - acc: 0.982 - ETA: 1s - loss: 0.0664 - acc: 0.981 - ETA: 1s - loss: 0.0659 - acc: 0.982 - ETA: 1s - loss: 0.0666 - acc: 0.982 - ETA: 1s - loss: 0.0671 - acc: 0.982 - ETA: 1s - loss: 0.0669 - acc: 0.982 - ETA: 1s - loss: 0.0665 - acc: 0.982 - ETA: 1s - loss: 0.0662 - acc: 0.982 - ETA: 1s - loss: 0.0659 - acc: 0.982 - ETA: 1s - loss: 0.0657 - acc: 0.982 - ETA: 1s - loss: 0.0656 - acc: 0.982 - ETA: 1s - loss: 0.0659 - acc: 0.982 - ETA: 1s - loss: 0.0659 - acc: 0.982 - ETA: 1s - loss: 0.0660 - acc: 0.982 - ETA: 1s - loss: 0.0666 - acc: 0.982 - ETA: 1s - loss: 0.0662 - acc: 0.982 - ETA: 1s - loss: 0.0657 - acc: 0.982 - ETA: 1s - loss: 0.0653 - acc: 0.982 - ETA: 1s - loss: 0.0648 - acc: 0.982 - ETA: 0s - loss: 0.0649 - acc: 0.982 - ETA: 0s - loss: 0.0646 - acc: 0.983 - ETA: 0s - loss: 0.0658 - acc: 0.983 - ETA: 0s - loss: 0.0654 - acc: 0.983 - ETA: 0s - loss: 0.0652 - acc: 0.983 - ETA: 0s - loss: 0.0648 - acc: 0.983 - ETA: 0s - loss: 0.0648 - acc: 0.983 - ETA: 0s - loss: 0.0651 - acc: 0.983 - ETA: 0s - loss: 0.0647 - acc: 0.983 - ETA: 0s - loss: 0.0651 - acc: 0.983 - ETA: 0s - loss: 0.0652 - acc: 0.983 - ETA: 0s - loss: 0.0653 - acc: 0.983 - ETA: 0s - loss: 0.0651 - acc: 0.983 - ETA: 0s - loss: 0.0662 - acc: 0.982 - ETA: 0s - loss: 0.0658 - acc: 0.982 - ETA: 0s - loss: 0.0654 - acc: 0.9829Epoch 00006: val_loss did not improve 6680/6680 [==============================] - 7s - loss: 0.0659 - acc: 0.9826 - val_loss: 0.7379 - val_acc: 0.8120 Epoch 8/20 6620/6680 [============================>.] - ETA: 8s - loss: 0.0019 - acc: 1.000 - ETA: 10s - loss: 0.0317 - acc: 1.00 - ETA: 8s - loss: 0.0223 - acc: 1.0000 - ETA: 8s - loss: 0.0256 - acc: 0.993 - ETA: 10s - loss: 0.0251 - acc: 0.99 - ETA: 10s - loss: 0.0254 - acc: 0.99 - ETA: 11s - loss: 0.0241 - acc: 0.99 - ETA: 10s - loss: 0.0241 - acc: 0.99 - ETA: 10s - loss: 0.0316 - acc: 0.99 - ETA: 9s - loss: 0.0293 - acc: 0.9950 - ETA: 8s - loss: 0.0278 - acc: 0.995 - ETA: 8s - loss: 0.0265 - acc: 0.996 - ETA: 8s - loss: 0.0254 - acc: 0.996 - ETA: 7s - loss: 0.0262 - acc: 0.995 - ETA: 7s - loss: 0.0254 - acc: 0.995 - ETA: 7s - loss: 0.0254 - acc: 0.996 - ETA: 6s - loss: 0.0258 - acc: 0.995 - ETA: 6s - loss: 0.0246 - acc: 0.995 - ETA: 6s - loss: 0.0245 - acc: 0.994 - ETA: 6s - loss: 0.0297 - acc: 0.993 - ETA: 6s - loss: 0.0297 - acc: 0.992 - ETA: 6s - loss: 0.0317 - acc: 0.992 - ETA: 6s - loss: 0.0311 - acc: 0.992 - ETA: 5s - loss: 0.0309 - acc: 0.992 - ETA: 5s - loss: 0.0300 - acc: 0.993 - ETA: 5s - loss: 0.0294 - acc: 0.993 - ETA: 5s - loss: 0.0289 - acc: 0.993 - ETA: 5s - loss: 0.0287 - acc: 0.993 - ETA: 5s - loss: 0.0336 - acc: 0.991 - ETA: 5s - loss: 0.0341 - acc: 0.990 - ETA: 5s - loss: 0.0335 - acc: 0.991 - ETA: 5s - loss: 0.0346 - acc: 0.990 - ETA: 5s - loss: 0.0380 - acc: 0.990 - ETA: 5s - loss: 0.0404 - acc: 0.989 - ETA: 5s - loss: 0.0405 - acc: 0.988 - ETA: 4s - loss: 0.0411 - acc: 0.988 - ETA: 4s - loss: 0.0416 - acc: 0.988 - ETA: 4s - loss: 0.0415 - acc: 0.988 - ETA: 4s - loss: 0.0415 - acc: 0.987 - ETA: 4s - loss: 0.0412 - acc: 0.987 - ETA: 5s - loss: 0.0410 - acc: 0.987 - ETA: 5s - loss: 0.0406 - acc: 0.988 - ETA: 5s - loss: 0.0402 - acc: 0.988 - ETA: 5s - loss: 0.0398 - acc: 0.988 - ETA: 5s - loss: 0.0403 - acc: 0.988 - ETA: 4s - loss: 0.0402 - acc: 0.988 - ETA: 4s - loss: 0.0398 - acc: 0.988 - ETA: 4s - loss: 0.0416 - acc: 0.987 - ETA: 4s - loss: 0.0410 - acc: 0.988 - ETA: 4s - loss: 0.0430 - acc: 0.987 - ETA: 4s - loss: 0.0422 - acc: 0.987 - ETA: 4s - loss: 0.0435 - acc: 0.987 - ETA: 4s - loss: 0.0450 - acc: 0.986 - ETA: 4s - loss: 0.0448 - acc: 0.986 - ETA: 4s - loss: 0.0443 - acc: 0.986 - ETA: 4s - loss: 0.0441 - acc: 0.987 - ETA: 3s - loss: 0.0445 - acc: 0.987 - ETA: 3s - loss: 0.0448 - acc: 0.986 - ETA: 3s - loss: 0.0445 - acc: 0.987 - ETA: 3s - loss: 0.0442 - acc: 0.987 - ETA: 3s - loss: 0.0458 - acc: 0.986 - ETA: 3s - loss: 0.0457 - acc: 0.986 - ETA: 3s - loss: 0.0452 - acc: 0.986 - ETA: 3s - loss: 0.0449 - acc: 0.986 - ETA: 3s - loss: 0.0444 - acc: 0.987 - ETA: 3s - loss: 0.0443 - acc: 0.987 - ETA: 3s - loss: 0.0447 - acc: 0.987 - ETA: 3s - loss: 0.0444 - acc: 0.987 - ETA: 3s - loss: 0.0441 - acc: 0.987 - ETA: 3s - loss: 0.0437 - acc: 0.987 - ETA: 2s - loss: 0.0438 - acc: 0.987 - ETA: 2s - loss: 0.0439 - acc: 0.987 - ETA: 2s - loss: 0.0439 - acc: 0.987 - ETA: 2s - loss: 0.0439 - acc: 0.987 - ETA: 2s - loss: 0.0438 - acc: 0.987 - ETA: 2s - loss: 0.0436 - acc: 0.987 - ETA: 2s - loss: 0.0435 - acc: 0.987 - ETA: 2s - loss: 0.0440 - acc: 0.987 - ETA: 2s - loss: 0.0438 - acc: 0.987 - ETA: 2s - loss: 0.0441 - acc: 0.987 - ETA: 2s - loss: 0.0437 - acc: 0.987 - ETA: 2s - loss: 0.0434 - acc: 0.987 - ETA: 2s - loss: 0.0432 - acc: 0.988 - ETA: 2s - loss: 0.0430 - acc: 0.988 - ETA: 2s - loss: 0.0424 - acc: 0.988 - ETA: 2s - loss: 0.0425 - acc: 0.988 - ETA: 2s - loss: 0.0442 - acc: 0.987 - ETA: 2s - loss: 0.0441 - acc: 0.987 - ETA: 1s - loss: 0.0447 - acc: 0.987 - ETA: 1s - loss: 0.0446 - acc: 0.987 - ETA: 1s - loss: 0.0457 - acc: 0.987 - ETA: 1s - loss: 0.0454 - acc: 0.987 - ETA: 1s - loss: 0.0454 - acc: 0.987 - ETA: 1s - loss: 0.0463 - acc: 0.986 - ETA: 1s - loss: 0.0459 - acc: 0.986 - ETA: 1s - loss: 0.0457 - acc: 0.986 - ETA: 1s - loss: 0.0453 - acc: 0.986 - ETA: 1s - loss: 0.0451 - acc: 0.986 - ETA: 1s - loss: 0.0452 - acc: 0.986 - ETA: 1s - loss: 0.0461 - acc: 0.986 - ETA: 1s - loss: 0.0467 - acc: 0.986 - ETA: 0s - loss: 0.0469 - acc: 0.986 - ETA: 0s - loss: 0.0468 - acc: 0.986 - ETA: 0s - loss: 0.0466 - acc: 0.986 - ETA: 0s - loss: 0.0462 - acc: 0.986 - ETA: 0s - loss: 0.0461 - acc: 0.986 - ETA: 0s - loss: 0.0458 - acc: 0.986 - ETA: 0s - loss: 0.0457 - acc: 0.987 - ETA: 0s - loss: 0.0457 - acc: 0.986 - ETA: 0s - loss: 0.0456 - acc: 0.986 - ETA: 0s - loss: 0.0455 - acc: 0.987 - ETA: 0s - loss: 0.0453 - acc: 0.987 - ETA: 0s - loss: 0.0454 - acc: 0.987 - ETA: 0s - loss: 0.0452 - acc: 0.987 - ETA: 0s - loss: 0.0475 - acc: 0.986 - ETA: 0s - loss: 0.0472 - acc: 0.986 - ETA: 0s - loss: 0.0471 - acc: 0.987 - ETA: 0s - loss: 0.0472 - acc: 0.986 - ETA: 0s - loss: 0.0481 - acc: 0.986 - ETA: 0s - loss: 0.0491 - acc: 0.986 - ETA: 0s - loss: 0.0489 - acc: 0.9863Epoch 00007: val_loss did not improve 6680/6680 [==============================] - 7s - loss: 0.0490 - acc: 0.9861 - val_loss: 0.7199 - val_acc: 0.8180 Epoch 9/20 6640/6680 [============================>.] - ETA: 7s - loss: 0.0040 - acc: 1.000 - ETA: 5s - loss: 0.0087 - acc: 1.000 - ETA: 5s - loss: 0.0095 - acc: 1.000 - ETA: 5s - loss: 0.0124 - acc: 1.000 - ETA: 5s - loss: 0.0128 - acc: 1.000 - ETA: 5s - loss: 0.0132 - acc: 1.000 - ETA: 5s - loss: 0.0135 - acc: 1.000 - ETA: 5s - loss: 0.0132 - acc: 1.000 - ETA: 5s - loss: 0.0129 - acc: 1.000 - ETA: 5s - loss: 0.0159 - acc: 0.998 - ETA: 5s - loss: 0.0206 - acc: 0.997 - ETA: 5s - loss: 0.0196 - acc: 0.997 - ETA: 5s - loss: 0.0200 - acc: 0.996 - ETA: 5s - loss: 0.0224 - acc: 0.995 - ETA: 5s - loss: 0.0220 - acc: 0.995 - ETA: 5s - loss: 0.0216 - acc: 0.995 - ETA: 5s - loss: 0.0212 - acc: 0.995 - ETA: 5s - loss: 0.0258 - acc: 0.993 - ETA: 5s - loss: 0.0250 - acc: 0.993 - ETA: 5s - loss: 0.0249 - acc: 0.993 - ETA: 5s - loss: 0.0281 - acc: 0.992 - ETA: 5s - loss: 0.0277 - acc: 0.992 - ETA: 5s - loss: 0.0278 - acc: 0.991 - ETA: 5s - loss: 0.0269 - acc: 0.992 - ETA: 5s - loss: 0.0265 - acc: 0.992 - ETA: 5s - loss: 0.0285 - acc: 0.991 - ETA: 5s - loss: 0.0296 - acc: 0.991 - ETA: 5s - loss: 0.0296 - acc: 0.990 - ETA: 5s - loss: 0.0303 - acc: 0.989 - ETA: 5s - loss: 0.0298 - acc: 0.990 - ETA: 5s - loss: 0.0302 - acc: 0.989 - ETA: 5s - loss: 0.0301 - acc: 0.989 - ETA: 4s - loss: 0.0304 - acc: 0.989 - ETA: 4s - loss: 0.0300 - acc: 0.989 - ETA: 4s - loss: 0.0324 - acc: 0.988 - ETA: 4s - loss: 0.0330 - acc: 0.987 - ETA: 4s - loss: 0.0323 - acc: 0.988 - ETA: 4s - loss: 0.0317 - acc: 0.988 - ETA: 4s - loss: 0.0314 - acc: 0.988 - ETA: 4s - loss: 0.0311 - acc: 0.988 - ETA: 4s - loss: 0.0309 - acc: 0.989 - ETA: 4s - loss: 0.0308 - acc: 0.989 - ETA: 4s - loss: 0.0311 - acc: 0.989 - ETA: 4s - loss: 0.0311 - acc: 0.989 - ETA: 3s - loss: 0.0306 - acc: 0.989 - ETA: 3s - loss: 0.0303 - acc: 0.989 - ETA: 3s - loss: 0.0299 - acc: 0.990 - ETA: 3s - loss: 0.0296 - acc: 0.990 - ETA: 3s - loss: 0.0295 - acc: 0.990 - ETA: 3s - loss: 0.0293 - acc: 0.990 - ETA: 3s - loss: 0.0292 - acc: 0.990 - ETA: 3s - loss: 0.0293 - acc: 0.990 - ETA: 3s - loss: 0.0295 - acc: 0.989 - ETA: 3s - loss: 0.0302 - acc: 0.989 - ETA: 3s - loss: 0.0306 - acc: 0.989 - ETA: 3s - loss: 0.0308 - acc: 0.989 - ETA: 3s - loss: 0.0305 - acc: 0.989 - ETA: 3s - loss: 0.0301 - acc: 0.989 - ETA: 3s - loss: 0.0301 - acc: 0.989 - ETA: 3s - loss: 0.0302 - acc: 0.989 - ETA: 3s - loss: 0.0309 - acc: 0.989 - ETA: 3s - loss: 0.0307 - acc: 0.989 - ETA: 3s - loss: 0.0307 - acc: 0.989 - ETA: 3s - loss: 0.0306 - acc: 0.989 - ETA: 3s - loss: 0.0305 - acc: 0.989 - ETA: 3s - loss: 0.0305 - acc: 0.989 - ETA: 3s - loss: 0.0310 - acc: 0.989 - ETA: 3s - loss: 0.0327 - acc: 0.989 - ETA: 2s - loss: 0.0323 - acc: 0.989 - ETA: 2s - loss: 0.0332 - acc: 0.989 - ETA: 2s - loss: 0.0333 - acc: 0.989 - ETA: 2s - loss: 0.0331 - acc: 0.989 - ETA: 2s - loss: 0.0330 - acc: 0.989 - ETA: 2s - loss: 0.0331 - acc: 0.989 - ETA: 2s - loss: 0.0328 - acc: 0.989 - ETA: 2s - loss: 0.0328 - acc: 0.989 - ETA: 2s - loss: 0.0336 - acc: 0.989 - ETA: 2s - loss: 0.0334 - acc: 0.989 - ETA: 2s - loss: 0.0339 - acc: 0.989 - ETA: 2s - loss: 0.0344 - acc: 0.989 - ETA: 2s - loss: 0.0344 - acc: 0.989 - ETA: 2s - loss: 0.0345 - acc: 0.989 - ETA: 1s - loss: 0.0348 - acc: 0.988 - ETA: 1s - loss: 0.0347 - acc: 0.988 - ETA: 1s - loss: 0.0352 - acc: 0.988 - ETA: 1s - loss: 0.0358 - acc: 0.988 - ETA: 1s - loss: 0.0358 - acc: 0.988 - ETA: 1s - loss: 0.0356 - acc: 0.988 - ETA: 1s - loss: 0.0356 - acc: 0.988 - ETA: 1s - loss: 0.0354 - acc: 0.988 - ETA: 1s - loss: 0.0353 - acc: 0.988 - ETA: 1s - loss: 0.0350 - acc: 0.988 - ETA: 1s - loss: 0.0351 - acc: 0.988 - ETA: 1s - loss: 0.0352 - acc: 0.988 - ETA: 1s - loss: 0.0351 - acc: 0.988 - ETA: 1s - loss: 0.0352 - acc: 0.988 - ETA: 1s - loss: 0.0349 - acc: 0.988 - ETA: 1s - loss: 0.0347 - acc: 0.988 - ETA: 1s - loss: 0.0345 - acc: 0.988 - ETA: 1s - loss: 0.0345 - acc: 0.989 - ETA: 1s - loss: 0.0341 - acc: 0.989 - ETA: 1s - loss: 0.0381 - acc: 0.988 - ETA: 0s - loss: 0.0382 - acc: 0.988 - ETA: 0s - loss: 0.0382 - acc: 0.988 - ETA: 0s - loss: 0.0382 - acc: 0.988 - ETA: 0s - loss: 0.0380 - acc: 0.988 - ETA: 0s - loss: 0.0378 - acc: 0.988 - ETA: 0s - loss: 0.0377 - acc: 0.988 - ETA: 0s - loss: 0.0378 - acc: 0.988 - ETA: 0s - loss: 0.0377 - acc: 0.988 - ETA: 0s - loss: 0.0379 - acc: 0.988 - ETA: 0s - loss: 0.0380 - acc: 0.988 - ETA: 0s - loss: 0.0377 - acc: 0.988 - ETA: 0s - loss: 0.0377 - acc: 0.988 - ETA: 0s - loss: 0.0379 - acc: 0.988 - ETA: 0s - loss: 0.0377 - acc: 0.988 - ETA: 0s - loss: 0.0375 - acc: 0.988 - ETA: 0s - loss: 0.0373 - acc: 0.988 - ETA: 0s - loss: 0.0372 - acc: 0.988 - ETA: 0s - loss: 0.0370 - acc: 0.9889Epoch 00008: val_loss did not improve 6680/6680 [==============================] - 7s - loss: 0.0370 - acc: 0.9889 - val_loss: 0.7041 - val_acc: 0.8287 Epoch 10/20 6640/6680 [============================>.] - ETA: 8s - loss: 0.0847 - acc: 0.950 - ETA: 6s - loss: 0.0356 - acc: 0.987 - ETA: 6s - loss: 0.0268 - acc: 0.992 - ETA: 6s - loss: 0.0274 - acc: 0.994 - ETA: 6s - loss: 0.0268 - acc: 0.995 - ETA: 6s - loss: 0.0232 - acc: 0.996 - ETA: 6s - loss: 0.0231 - acc: 0.994 - ETA: 6s - loss: 0.0362 - acc: 0.992 - ETA: 5s - loss: 0.0323 - acc: 0.993 - ETA: 5s - loss: 0.0318 - acc: 0.992 - ETA: 5s - loss: 0.0292 - acc: 0.993 - ETA: 5s - loss: 0.0272 - acc: 0.993 - ETA: 5s - loss: 0.0282 - acc: 0.991 - ETA: 5s - loss: 0.0289 - acc: 0.991 - ETA: 5s - loss: 0.0279 - acc: 0.991 - ETA: 5s - loss: 0.0264 - acc: 0.992 - ETA: 5s - loss: 0.0257 - acc: 0.992 - ETA: 5s - loss: 0.0242 - acc: 0.993 - ETA: 5s - loss: 0.0227 - acc: 0.993 - ETA: 4s - loss: 0.0222 - acc: 0.994 - ETA: 4s - loss: 0.0225 - acc: 0.993 - ETA: 4s - loss: 0.0218 - acc: 0.994 - ETA: 4s - loss: 0.0212 - acc: 0.994 - ETA: 4s - loss: 0.0226 - acc: 0.993 - ETA: 4s - loss: 0.0224 - acc: 0.994 - ETA: 4s - loss: 0.0220 - acc: 0.994 - ETA: 4s - loss: 0.0219 - acc: 0.994 - ETA: 4s - loss: 0.0217 - acc: 0.994 - ETA: 4s - loss: 0.0211 - acc: 0.994 - ETA: 4s - loss: 0.0207 - acc: 0.994 - ETA: 4s - loss: 0.0202 - acc: 0.994 - ETA: 4s - loss: 0.0205 - acc: 0.995 - ETA: 4s - loss: 0.0204 - acc: 0.995 - ETA: 4s - loss: 0.0208 - acc: 0.995 - ETA: 4s - loss: 0.0205 - acc: 0.995 - ETA: 4s - loss: 0.0205 - acc: 0.995 - ETA: 4s - loss: 0.0201 - acc: 0.995 - ETA: 4s - loss: 0.0202 - acc: 0.995 - ETA: 4s - loss: 0.0215 - acc: 0.995 - ETA: 4s - loss: 0.0221 - acc: 0.995 - ETA: 4s - loss: 0.0217 - acc: 0.995 - ETA: 4s - loss: 0.0217 - acc: 0.995 - ETA: 4s - loss: 0.0214 - acc: 0.995 - ETA: 4s - loss: 0.0211 - acc: 0.995 - ETA: 4s - loss: 0.0207 - acc: 0.995 - ETA: 4s - loss: 0.0206 - acc: 0.995 - ETA: 4s - loss: 0.0214 - acc: 0.995 - ETA: 3s - loss: 0.0213 - acc: 0.995 - ETA: 3s - loss: 0.0211 - acc: 0.995 - ETA: 3s - loss: 0.0209 - acc: 0.995 - ETA: 3s - loss: 0.0223 - acc: 0.995 - ETA: 3s - loss: 0.0230 - acc: 0.995 - ETA: 3s - loss: 0.0231 - acc: 0.995 - ETA: 3s - loss: 0.0228 - acc: 0.995 - ETA: 3s - loss: 0.0230 - acc: 0.995 - ETA: 3s - loss: 0.0230 - acc: 0.995 - ETA: 3s - loss: 0.0232 - acc: 0.994 - ETA: 3s - loss: 0.0232 - acc: 0.994 - ETA: 3s - loss: 0.0235 - acc: 0.994 - ETA: 3s - loss: 0.0240 - acc: 0.994 - ETA: 2s - loss: 0.0241 - acc: 0.994 - ETA: 2s - loss: 0.0242 - acc: 0.994 - ETA: 2s - loss: 0.0239 - acc: 0.994 - ETA: 2s - loss: 0.0236 - acc: 0.994 - ETA: 2s - loss: 0.0242 - acc: 0.993 - ETA: 2s - loss: 0.0240 - acc: 0.994 - ETA: 2s - loss: 0.0239 - acc: 0.994 - ETA: 2s - loss: 0.0237 - acc: 0.994 - ETA: 2s - loss: 0.0237 - acc: 0.994 - ETA: 2s - loss: 0.0236 - acc: 0.994 - ETA: 2s - loss: 0.0234 - acc: 0.994 - ETA: 2s - loss: 0.0234 - acc: 0.994 - ETA: 2s - loss: 0.0233 - acc: 0.994 - ETA: 2s - loss: 0.0234 - acc: 0.994 - ETA: 2s - loss: 0.0232 - acc: 0.994 - ETA: 2s - loss: 0.0230 - acc: 0.994 - ETA: 2s - loss: 0.0229 - acc: 0.994 - ETA: 2s - loss: 0.0230 - acc: 0.994 - ETA: 2s - loss: 0.0234 - acc: 0.994 - ETA: 1s - loss: 0.0233 - acc: 0.994 - ETA: 1s - loss: 0.0234 - acc: 0.994 - ETA: 1s - loss: 0.0233 - acc: 0.994 - ETA: 1s - loss: 0.0231 - acc: 0.994 - ETA: 1s - loss: 0.0229 - acc: 0.994 - ETA: 1s - loss: 0.0235 - acc: 0.994 - ETA: 1s - loss: 0.0235 - acc: 0.994 - ETA: 1s - loss: 0.0234 - acc: 0.994 - ETA: 1s - loss: 0.0232 - acc: 0.994 - ETA: 1s - loss: 0.0230 - acc: 0.994 - ETA: 1s - loss: 0.0228 - acc: 0.994 - ETA: 1s - loss: 0.0231 - acc: 0.994 - ETA: 1s - loss: 0.0232 - acc: 0.994 - ETA: 0s - loss: 0.0231 - acc: 0.994 - ETA: 0s - loss: 0.0230 - acc: 0.994 - ETA: 0s - loss: 0.0231 - acc: 0.994 - ETA: 0s - loss: 0.0230 - acc: 0.994 - ETA: 0s - loss: 0.0231 - acc: 0.994 - ETA: 0s - loss: 0.0232 - acc: 0.994 - ETA: 0s - loss: 0.0234 - acc: 0.994 - ETA: 0s - loss: 0.0235 - acc: 0.993 - ETA: 0s - loss: 0.0236 - acc: 0.993 - ETA: 0s - loss: 0.0236 - acc: 0.993 - ETA: 0s - loss: 0.0237 - acc: 0.993 - ETA: 0s - loss: 0.0239 - acc: 0.993 - ETA: 0s - loss: 0.0251 - acc: 0.993 - ETA: 0s - loss: 0.0252 - acc: 0.993 - ETA: 0s - loss: 0.0261 - acc: 0.992 - ETA: 0s - loss: 0.0261 - acc: 0.992 - ETA: 0s - loss: 0.0260 - acc: 0.992 - ETA: 0s - loss: 0.0262 - acc: 0.992 - ETA: 0s - loss: 0.0261 - acc: 0.992 - ETA: 0s - loss: 0.0265 - acc: 0.9928Epoch 00009: val_loss did not improve 6680/6680 [==============================] - 6s - loss: 0.0265 - acc: 0.9928 - val_loss: 0.7062 - val_acc: 0.8216 Epoch 11/20 6660/6680 [============================>.] - ETA: 7s - loss: 0.0068 - acc: 1.000 - ETA: 5s - loss: 0.0076 - acc: 1.000 - ETA: 5s - loss: 0.0139 - acc: 0.994 - ETA: 5s - loss: 0.0119 - acc: 0.995 - ETA: 5s - loss: 0.0113 - acc: 0.996 - ETA: 5s - loss: 0.0099 - acc: 0.997 - ETA: 4s - loss: 0.0098 - acc: 0.997 - ETA: 4s - loss: 0.0092 - acc: 0.998 - ETA: 4s - loss: 0.0098 - acc: 0.998 - ETA: 4s - loss: 0.0094 - acc: 0.998 - ETA: 4s - loss: 0.0095 - acc: 0.998 - ETA: 4s - loss: 0.0103 - acc: 0.998 - ETA: 4s - loss: 0.0101 - acc: 0.998 - ETA: 4s - loss: 0.0119 - acc: 0.997 - ETA: 4s - loss: 0.0130 - acc: 0.996 - ETA: 4s - loss: 0.0152 - acc: 0.996 - ETA: 5s - loss: 0.0150 - acc: 0.996 - ETA: 4s - loss: 0.0146 - acc: 0.996 - ETA: 5s - loss: 0.0145 - acc: 0.996 - ETA: 5s - loss: 0.0146 - acc: 0.996 - ETA: 5s - loss: 0.0142 - acc: 0.996 - ETA: 5s - loss: 0.0152 - acc: 0.996 - ETA: 5s - loss: 0.0155 - acc: 0.996 - ETA: 5s - loss: 0.0153 - acc: 0.996 - ETA: 5s - loss: 0.0179 - acc: 0.995 - ETA: 5s - loss: 0.0173 - acc: 0.995 - ETA: 5s - loss: 0.0167 - acc: 0.995 - ETA: 4s - loss: 0.0166 - acc: 0.995 - ETA: 4s - loss: 0.0206 - acc: 0.994 - ETA: 4s - loss: 0.0203 - acc: 0.994 - ETA: 4s - loss: 0.0198 - acc: 0.995 - ETA: 4s - loss: 0.0195 - acc: 0.995 - ETA: 4s - loss: 0.0191 - acc: 0.995 - ETA: 4s - loss: 0.0187 - acc: 0.995 - ETA: 4s - loss: 0.0182 - acc: 0.995 - ETA: 4s - loss: 0.0222 - acc: 0.994 - ETA: 4s - loss: 0.0221 - acc: 0.994 - ETA: 4s - loss: 0.0235 - acc: 0.993 - ETA: 4s - loss: 0.0236 - acc: 0.993 - ETA: 4s - loss: 0.0232 - acc: 0.993 - ETA: 3s - loss: 0.0228 - acc: 0.993 - ETA: 3s - loss: 0.0226 - acc: 0.994 - ETA: 3s - loss: 0.0220 - acc: 0.994 - ETA: 3s - loss: 0.0219 - acc: 0.994 - ETA: 3s - loss: 0.0219 - acc: 0.994 - ETA: 3s - loss: 0.0215 - acc: 0.994 - ETA: 3s - loss: 0.0216 - acc: 0.993 - ETA: 3s - loss: 0.0215 - acc: 0.994 - ETA: 3s - loss: 0.0217 - acc: 0.993 - ETA: 3s - loss: 0.0218 - acc: 0.993 - ETA: 3s - loss: 0.0216 - acc: 0.994 - ETA: 3s - loss: 0.0213 - acc: 0.994 - ETA: 3s - loss: 0.0230 - acc: 0.993 - ETA: 3s - loss: 0.0226 - acc: 0.994 - ETA: 3s - loss: 0.0223 - acc: 0.994 - ETA: 3s - loss: 0.0219 - acc: 0.994 - ETA: 3s - loss: 0.0222 - acc: 0.994 - ETA: 3s - loss: 0.0225 - acc: 0.994 - ETA: 2s - loss: 0.0225 - acc: 0.993 - ETA: 2s - loss: 0.0225 - acc: 0.993 - ETA: 2s - loss: 0.0222 - acc: 0.994 - ETA: 2s - loss: 0.0219 - acc: 0.994 - ETA: 2s - loss: 0.0216 - acc: 0.994 - ETA: 2s - loss: 0.0214 - acc: 0.994 - ETA: 2s - loss: 0.0213 - acc: 0.994 - ETA: 2s - loss: 0.0212 - acc: 0.994 - ETA: 2s - loss: 0.0213 - acc: 0.994 - ETA: 2s - loss: 0.0211 - acc: 0.994 - ETA: 2s - loss: 0.0209 - acc: 0.994 - ETA: 2s - loss: 0.0207 - acc: 0.994 - ETA: 2s - loss: 0.0209 - acc: 0.994 - ETA: 2s - loss: 0.0210 - acc: 0.994 - ETA: 2s - loss: 0.0221 - acc: 0.994 - ETA: 2s - loss: 0.0220 - acc: 0.994 - ETA: 1s - loss: 0.0218 - acc: 0.994 - ETA: 1s - loss: 0.0216 - acc: 0.994 - ETA: 1s - loss: 0.0218 - acc: 0.994 - ETA: 1s - loss: 0.0218 - acc: 0.994 - ETA: 1s - loss: 0.0216 - acc: 0.994 - ETA: 1s - loss: 0.0215 - acc: 0.994 - ETA: 1s - loss: 0.0216 - acc: 0.994 - ETA: 1s - loss: 0.0215 - acc: 0.994 - ETA: 1s - loss: 0.0215 - acc: 0.994 - ETA: 1s - loss: 0.0213 - acc: 0.994 - ETA: 1s - loss: 0.0216 - acc: 0.994 - ETA: 1s - loss: 0.0217 - acc: 0.994 - ETA: 1s - loss: 0.0214 - acc: 0.994 - ETA: 1s - loss: 0.0213 - acc: 0.994 - ETA: 1s - loss: 0.0212 - acc: 0.994 - ETA: 0s - loss: 0.0219 - acc: 0.994 - ETA: 0s - loss: 0.0237 - acc: 0.993 - ETA: 0s - loss: 0.0237 - acc: 0.994 - ETA: 0s - loss: 0.0238 - acc: 0.993 - ETA: 0s - loss: 0.0236 - acc: 0.993 - ETA: 0s - loss: 0.0235 - acc: 0.994 - ETA: 0s - loss: 0.0233 - acc: 0.994 - ETA: 0s - loss: 0.0231 - acc: 0.994 - ETA: 0s - loss: 0.0233 - acc: 0.994 - ETA: 0s - loss: 0.0238 - acc: 0.993 - ETA: 0s - loss: 0.0238 - acc: 0.994 - ETA: 0s - loss: 0.0238 - acc: 0.994 - ETA: 0s - loss: 0.0238 - acc: 0.994 - ETA: 0s - loss: 0.0237 - acc: 0.994 - ETA: 0s - loss: 0.0236 - acc: 0.994 - ETA: 0s - loss: 0.0235 - acc: 0.994 - ETA: 0s - loss: 0.0234 - acc: 0.994 - ETA: 0s - loss: 0.0233 - acc: 0.9943Epoch 00010: val_loss did not improve 6680/6680 [==============================] - 6s - loss: 0.0234 - acc: 0.9942 - val_loss: 0.8153 - val_acc: 0.8144 Epoch 12/20 6660/6680 [============================>.] - ETA: 5s - loss: 3.0918e-04 - acc: 1.000 - ETA: 4s - loss: 0.0034 - acc: 1.0000 - ETA: 4s - loss: 0.0037 - acc: 1.000 - ETA: 4s - loss: 0.0086 - acc: 1.000 - ETA: 4s - loss: 0.0091 - acc: 1.000 - ETA: 5s - loss: 0.0087 - acc: 1.000 - ETA: 5s - loss: 0.0076 - acc: 1.000 - ETA: 5s - loss: 0.0070 - acc: 1.000 - ETA: 5s - loss: 0.0068 - acc: 1.000 - ETA: 5s - loss: 0.0064 - acc: 1.000 - ETA: 5s - loss: 0.0081 - acc: 1.000 - ETA: 5s - loss: 0.0076 - acc: 1.000 - ETA: 5s - loss: 0.0073 - acc: 1.000 - ETA: 4s - loss: 0.0164 - acc: 0.998 - ETA: 4s - loss: 0.0152 - acc: 0.998 - ETA: 4s - loss: 0.0170 - acc: 0.998 - ETA: 4s - loss: 0.0170 - acc: 0.997 - ETA: 4s - loss: 0.0164 - acc: 0.997 - ETA: 4s - loss: 0.0158 - acc: 0.997 - ETA: 4s - loss: 0.0162 - acc: 0.997 - ETA: 4s - loss: 0.0158 - acc: 0.997 - ETA: 4s - loss: 0.0153 - acc: 0.997 - ETA: 4s - loss: 0.0151 - acc: 0.998 - ETA: 4s - loss: 0.0145 - acc: 0.998 - ETA: 4s - loss: 0.0146 - acc: 0.998 - ETA: 4s - loss: 0.0141 - acc: 0.998 - ETA: 4s - loss: 0.0161 - acc: 0.997 - ETA: 3s - loss: 0.0155 - acc: 0.997 - ETA: 3s - loss: 0.0153 - acc: 0.998 - ETA: 3s - loss: 0.0150 - acc: 0.998 - ETA: 3s - loss: 0.0149 - acc: 0.998 - ETA: 3s - loss: 0.0146 - acc: 0.998 - ETA: 3s - loss: 0.0151 - acc: 0.997 - ETA: 3s - loss: 0.0147 - acc: 0.997 - ETA: 3s - loss: 0.0145 - acc: 0.997 - ETA: 3s - loss: 0.0141 - acc: 0.998 - ETA: 3s - loss: 0.0143 - acc: 0.997 - ETA: 3s - loss: 0.0141 - acc: 0.997 - ETA: 3s - loss: 0.0140 - acc: 0.997 - ETA: 3s - loss: 0.0157 - acc: 0.997 - ETA: 3s - loss: 0.0155 - acc: 0.997 - ETA: 3s - loss: 0.0158 - acc: 0.996 - ETA: 3s - loss: 0.0155 - acc: 0.997 - ETA: 2s - loss: 0.0167 - acc: 0.996 - ETA: 2s - loss: 0.0164 - acc: 0.996 - ETA: 2s - loss: 0.0162 - acc: 0.996 - ETA: 2s - loss: 0.0182 - acc: 0.996 - ETA: 2s - loss: 0.0182 - acc: 0.996 - ETA: 2s - loss: 0.0215 - acc: 0.995 - ETA: 2s - loss: 0.0212 - acc: 0.995 - ETA: 2s - loss: 0.0209 - acc: 0.995 - ETA: 2s - loss: 0.0207 - acc: 0.995 - ETA: 2s - loss: 0.0204 - acc: 0.995 - ETA: 2s - loss: 0.0205 - acc: 0.995 - ETA: 2s - loss: 0.0203 - acc: 0.995 - ETA: 2s - loss: 0.0201 - acc: 0.995 - ETA: 2s - loss: 0.0203 - acc: 0.995 - ETA: 2s - loss: 0.0201 - acc: 0.995 - ETA: 2s - loss: 0.0197 - acc: 0.995 - ETA: 2s - loss: 0.0195 - acc: 0.995 - ETA: 2s - loss: 0.0205 - acc: 0.995 - ETA: 2s - loss: 0.0202 - acc: 0.995 - ETA: 1s - loss: 0.0200 - acc: 0.995 - ETA: 1s - loss: 0.0198 - acc: 0.995 - ETA: 1s - loss: 0.0207 - acc: 0.995 - ETA: 1s - loss: 0.0205 - acc: 0.995 - ETA: 1s - loss: 0.0203 - acc: 0.995 - ETA: 1s - loss: 0.0201 - acc: 0.995 - ETA: 1s - loss: 0.0212 - acc: 0.995 - ETA: 1s - loss: 0.0214 - acc: 0.995 - ETA: 1s - loss: 0.0215 - acc: 0.994 - ETA: 1s - loss: 0.0214 - acc: 0.994 - ETA: 1s - loss: 0.0213 - acc: 0.994 - ETA: 1s - loss: 0.0211 - acc: 0.994 - ETA: 1s - loss: 0.0209 - acc: 0.995 - ETA: 1s - loss: 0.0208 - acc: 0.995 - ETA: 1s - loss: 0.0207 - acc: 0.995 - ETA: 1s - loss: 0.0205 - acc: 0.995 - ETA: 0s - loss: 0.0203 - acc: 0.995 - ETA: 0s - loss: 0.0208 - acc: 0.995 - ETA: 0s - loss: 0.0221 - acc: 0.994 - ETA: 0s - loss: 0.0220 - acc: 0.994 - ETA: 0s - loss: 0.0219 - acc: 0.995 - ETA: 0s - loss: 0.0218 - acc: 0.994 - ETA: 0s - loss: 0.0216 - acc: 0.994 - ETA: 0s - loss: 0.0214 - acc: 0.995 - ETA: 0s - loss: 0.0212 - acc: 0.995 - ETA: 0s - loss: 0.0212 - acc: 0.995 - ETA: 0s - loss: 0.0211 - acc: 0.995 - ETA: 0s - loss: 0.0214 - acc: 0.994 - ETA: 0s - loss: 0.0214 - acc: 0.995 - ETA: 0s - loss: 0.0212 - acc: 0.995 - ETA: 0s - loss: 0.0210 - acc: 0.995 - ETA: 0s - loss: 0.0209 - acc: 0.995 - ETA: 0s - loss: 0.0207 - acc: 0.9952Epoch 00011: val_loss did not improve 6680/6680 [==============================] - 5s - loss: 0.0207 - acc: 0.9952 - val_loss: 0.8267 - val_acc: 0.8192 Epoch 13/20 6620/6680 [============================>.] - ETA: 15s - loss: 0.0142 - acc: 1.00 - ETA: 7s - loss: 0.0053 - acc: 1.0000 - ETA: 7s - loss: 0.0047 - acc: 1.000 - ETA: 6s - loss: 0.0042 - acc: 1.000 - ETA: 6s - loss: 0.0045 - acc: 1.000 - ETA: 5s - loss: 0.0040 - acc: 1.000 - ETA: 5s - loss: 0.0041 - acc: 1.000 - ETA: 5s - loss: 0.0041 - acc: 1.000 - ETA: 5s - loss: 0.0040 - acc: 1.000 - ETA: 5s - loss: 0.0038 - acc: 1.000 - ETA: 4s - loss: 0.0050 - acc: 0.998 - ETA: 4s - loss: 0.0047 - acc: 0.998 - ETA: 4s - loss: 0.0067 - acc: 0.997 - ETA: 4s - loss: 0.0071 - acc: 0.997 - ETA: 4s - loss: 0.0086 - acc: 0.997 - ETA: 4s - loss: 0.0083 - acc: 0.997 - ETA: 4s - loss: 0.0084 - acc: 0.997 - ETA: 4s - loss: 0.0081 - acc: 0.997 - ETA: 4s - loss: 0.0078 - acc: 0.997 - ETA: 4s - loss: 0.0075 - acc: 0.997 - ETA: 4s - loss: 0.0073 - acc: 0.998 - ETA: 4s - loss: 0.0071 - acc: 0.998 - ETA: 3s - loss: 0.0076 - acc: 0.997 - ETA: 3s - loss: 0.0098 - acc: 0.996 - ETA: 3s - loss: 0.0096 - acc: 0.996 - ETA: 3s - loss: 0.0110 - acc: 0.996 - ETA: 3s - loss: 0.0111 - acc: 0.996 - ETA: 3s - loss: 0.0111 - acc: 0.996 - ETA: 3s - loss: 0.0108 - acc: 0.996 - ETA: 3s - loss: 0.0105 - acc: 0.996 - ETA: 3s - loss: 0.0105 - acc: 0.996 - ETA: 3s - loss: 0.0107 - acc: 0.997 - ETA: 3s - loss: 0.0107 - acc: 0.997 - ETA: 3s - loss: 0.0105 - acc: 0.997 - ETA: 3s - loss: 0.0105 - acc: 0.997 - ETA: 3s - loss: 0.0115 - acc: 0.997 - ETA: 2s - loss: 0.0144 - acc: 0.996 - ETA: 2s - loss: 0.0142 - acc: 0.996 - ETA: 2s - loss: 0.0152 - acc: 0.996 - ETA: 2s - loss: 0.0150 - acc: 0.996 - ETA: 2s - loss: 0.0160 - acc: 0.996 - ETA: 2s - loss: 0.0164 - acc: 0.996 - ETA: 2s - loss: 0.0161 - acc: 0.996 - ETA: 2s - loss: 0.0163 - acc: 0.996 - ETA: 2s - loss: 0.0161 - acc: 0.996 - ETA: 2s - loss: 0.0162 - acc: 0.996 - ETA: 2s - loss: 0.0161 - acc: 0.996 - ETA: 2s - loss: 0.0158 - acc: 0.996 - ETA: 2s - loss: 0.0157 - acc: 0.996 - ETA: 2s - loss: 0.0157 - acc: 0.996 - ETA: 2s - loss: 0.0154 - acc: 0.996 - ETA: 2s - loss: 0.0153 - acc: 0.996 - ETA: 2s - loss: 0.0150 - acc: 0.996 - ETA: 1s - loss: 0.0148 - acc: 0.996 - ETA: 1s - loss: 0.0155 - acc: 0.996 - ETA: 1s - loss: 0.0170 - acc: 0.995 - ETA: 1s - loss: 0.0168 - acc: 0.995 - ETA: 1s - loss: 0.0169 - acc: 0.995 - ETA: 1s - loss: 0.0171 - acc: 0.995 - ETA: 1s - loss: 0.0169 - acc: 0.995 - ETA: 1s - loss: 0.0171 - acc: 0.995 - ETA: 1s - loss: 0.0168 - acc: 0.995 - ETA: 1s - loss: 0.0173 - acc: 0.995 - ETA: 1s - loss: 0.0174 - acc: 0.995 - ETA: 1s - loss: 0.0171 - acc: 0.995 - ETA: 1s - loss: 0.0169 - acc: 0.995 - ETA: 1s - loss: 0.0167 - acc: 0.995 - ETA: 1s - loss: 0.0166 - acc: 0.995 - ETA: 1s - loss: 0.0166 - acc: 0.995 - ETA: 1s - loss: 0.0164 - acc: 0.995 - ETA: 0s - loss: 0.0163 - acc: 0.995 - ETA: 0s - loss: 0.0163 - acc: 0.995 - ETA: 0s - loss: 0.0161 - acc: 0.995 - ETA: 0s - loss: 0.0160 - acc: 0.995 - ETA: 0s - loss: 0.0158 - acc: 0.995 - ETA: 0s - loss: 0.0159 - acc: 0.995 - ETA: 0s - loss: 0.0164 - acc: 0.995 - ETA: 0s - loss: 0.0162 - acc: 0.995 - ETA: 0s - loss: 0.0161 - acc: 0.995 - ETA: 0s - loss: 0.0159 - acc: 0.995 - ETA: 0s - loss: 0.0161 - acc: 0.995 - ETA: 0s - loss: 0.0162 - acc: 0.995 - ETA: 0s - loss: 0.0160 - acc: 0.995 - ETA: 0s - loss: 0.0159 - acc: 0.995 - ETA: 0s - loss: 0.0158 - acc: 0.995 - ETA: 0s - loss: 0.0159 - acc: 0.995 - ETA: 0s - loss: 0.0157 - acc: 0.9956Epoch 00012: val_loss did not improve 6680/6680 [==============================] - 5s - loss: 0.0156 - acc: 0.9957 - val_loss: 0.7601 - val_acc: 0.8240 Epoch 14/20 6640/6680 [============================>.] - ETA: 16s - loss: 4.4875e-04 - acc: 1.00 - ETA: 6s - loss: 0.0017 - acc: 1.0000 - ETA: 6s - loss: 0.0019 - acc: 1.000 - ETA: 5s - loss: 0.0031 - acc: 1.000 - ETA: 5s - loss: 0.0046 - acc: 1.000 - ETA: 5s - loss: 0.0039 - acc: 1.000 - ETA: 4s - loss: 0.0040 - acc: 1.000 - ETA: 4s - loss: 0.0036 - acc: 1.000 - ETA: 4s - loss: 0.0035 - acc: 1.000 - ETA: 4s - loss: 0.0034 - acc: 1.000 - ETA: 4s - loss: 0.0032 - acc: 1.000 - ETA: 4s - loss: 0.0038 - acc: 1.000 - ETA: 4s - loss: 0.0036 - acc: 1.000 - ETA: 4s - loss: 0.0039 - acc: 1.000 - ETA: 4s - loss: 0.0041 - acc: 1.000 - ETA: 4s - loss: 0.0089 - acc: 0.999 - ETA: 4s - loss: 0.0090 - acc: 0.999 - ETA: 4s - loss: 0.0087 - acc: 0.999 - ETA: 4s - loss: 0.0085 - acc: 0.999 - ETA: 3s - loss: 0.0081 - acc: 0.999 - ETA: 3s - loss: 0.0079 - acc: 0.999 - ETA: 3s - loss: 0.0086 - acc: 0.998 - ETA: 3s - loss: 0.0084 - acc: 0.998 - ETA: 3s - loss: 0.0087 - acc: 0.998 - ETA: 3s - loss: 0.0084 - acc: 0.998 - ETA: 3s - loss: 0.0083 - acc: 0.998 - ETA: 3s - loss: 0.0082 - acc: 0.998 - ETA: 3s - loss: 0.0080 - acc: 0.998 - ETA: 3s - loss: 0.0081 - acc: 0.998 - ETA: 3s - loss: 0.0083 - acc: 0.998 - ETA: 3s - loss: 0.0081 - acc: 0.998 - ETA: 3s - loss: 0.0079 - acc: 0.998 - ETA: 3s - loss: 0.0079 - acc: 0.998 - ETA: 3s - loss: 0.0078 - acc: 0.998 - ETA: 3s - loss: 0.0076 - acc: 0.998 - ETA: 3s - loss: 0.0079 - acc: 0.998 - ETA: 3s - loss: 0.0077 - acc: 0.998 - ETA: 2s - loss: 0.0076 - acc: 0.998 - ETA: 2s - loss: 0.0082 - acc: 0.998 - ETA: 2s - loss: 0.0081 - acc: 0.998 - ETA: 2s - loss: 0.0081 - acc: 0.998 - ETA: 2s - loss: 0.0079 - acc: 0.998 - ETA: 2s - loss: 0.0079 - acc: 0.998 - ETA: 2s - loss: 0.0079 - acc: 0.998 - ETA: 2s - loss: 0.0079 - acc: 0.998 - ETA: 2s - loss: 0.0085 - acc: 0.998 - ETA: 2s - loss: 0.0085 - acc: 0.998 - ETA: 2s - loss: 0.0084 - acc: 0.998 - ETA: 2s - loss: 0.0083 - acc: 0.998 - ETA: 2s - loss: 0.0082 - acc: 0.998 - ETA: 2s - loss: 0.0083 - acc: 0.998 - ETA: 2s - loss: 0.0082 - acc: 0.998 - ETA: 2s - loss: 0.0081 - acc: 0.998 - ETA: 2s - loss: 0.0100 - acc: 0.997 - ETA: 2s - loss: 0.0099 - acc: 0.997 - ETA: 2s - loss: 0.0099 - acc: 0.998 - ETA: 2s - loss: 0.0098 - acc: 0.998 - ETA: 2s - loss: 0.0098 - acc: 0.998 - ETA: 2s - loss: 0.0097 - acc: 0.998 - ETA: 2s - loss: 0.0096 - acc: 0.998 - ETA: 1s - loss: 0.0096 - acc: 0.998 - ETA: 1s - loss: 0.0095 - acc: 0.998 - ETA: 1s - loss: 0.0094 - acc: 0.998 - ETA: 1s - loss: 0.0094 - acc: 0.998 - ETA: 1s - loss: 0.0093 - acc: 0.998 - ETA: 1s - loss: 0.0092 - acc: 0.998 - ETA: 1s - loss: 0.0099 - acc: 0.998 - ETA: 1s - loss: 0.0099 - acc: 0.998 - ETA: 1s - loss: 0.0098 - acc: 0.998 - ETA: 1s - loss: 0.0099 - acc: 0.997 - ETA: 1s - loss: 0.0099 - acc: 0.997 - ETA: 1s - loss: 0.0099 - acc: 0.998 - ETA: 1s - loss: 0.0098 - acc: 0.998 - ETA: 1s - loss: 0.0099 - acc: 0.997 - ETA: 1s - loss: 0.0114 - acc: 0.997 - ETA: 1s - loss: 0.0113 - acc: 0.997 - ETA: 1s - loss: 0.0112 - acc: 0.997 - ETA: 1s - loss: 0.0112 - acc: 0.997 - ETA: 1s - loss: 0.0110 - acc: 0.997 - ETA: 1s - loss: 0.0114 - acc: 0.997 - ETA: 0s - loss: 0.0115 - acc: 0.997 - ETA: 0s - loss: 0.0115 - acc: 0.997 - ETA: 0s - loss: 0.0114 - acc: 0.997 - ETA: 0s - loss: 0.0116 - acc: 0.997 - ETA: 0s - loss: 0.0117 - acc: 0.997 - ETA: 0s - loss: 0.0117 - acc: 0.997 - ETA: 0s - loss: 0.0117 - acc: 0.997 - ETA: 0s - loss: 0.0116 - acc: 0.997 - ETA: 0s - loss: 0.0115 - acc: 0.997 - ETA: 0s - loss: 0.0114 - acc: 0.997 - ETA: 0s - loss: 0.0115 - acc: 0.997 - ETA: 0s - loss: 0.0114 - acc: 0.997 - ETA: 0s - loss: 0.0115 - acc: 0.997 - ETA: 0s - loss: 0.0115 - acc: 0.997 - ETA: 0s - loss: 0.0114 - acc: 0.997 - ETA: 0s - loss: 0.0113 - acc: 0.997 - ETA: 0s - loss: 0.0112 - acc: 0.997 - ETA: 0s - loss: 0.0112 - acc: 0.997 - ETA: 0s - loss: 0.0112 - acc: 0.997 - ETA: 0s - loss: 0.0112 - acc: 0.9974Epoch 00013: val_loss did not improve 6680/6680 [==============================] - 5s - loss: 0.0111 - acc: 0.9975 - val_loss: 0.8287 - val_acc: 0.8192 Epoch 15/20 6620/6680 [============================>.] - ETA: 6s - loss: 0.0016 - acc: 1.000 - ETA: 5s - loss: 0.0011 - acc: 1.000 - ETA: 5s - loss: 0.0022 - acc: 1.000 - ETA: 5s - loss: 0.0024 - acc: 1.000 - ETA: 5s - loss: 0.0020 - acc: 1.000 - ETA: 5s - loss: 0.0020 - acc: 1.000 - ETA: 5s - loss: 0.0162 - acc: 0.997 - ETA: 5s - loss: 0.0147 - acc: 0.997 - ETA: 5s - loss: 0.0217 - acc: 0.996 - ETA: 5s - loss: 0.0197 - acc: 0.996 - ETA: 5s - loss: 0.0180 - acc: 0.997 - ETA: 5s - loss: 0.0167 - acc: 0.997 - ETA: 5s - loss: 0.0155 - acc: 0.997 - ETA: 5s - loss: 0.0144 - acc: 0.997 - ETA: 5s - loss: 0.0144 - acc: 0.997 - ETA: 5s - loss: 0.0133 - acc: 0.997 - ETA: 4s - loss: 0.0126 - acc: 0.998 - ETA: 5s - loss: 0.0122 - acc: 0.998 - ETA: 5s - loss: 0.0117 - acc: 0.998 - ETA: 5s - loss: 0.0113 - acc: 0.998 - ETA: 5s - loss: 0.0110 - acc: 0.998 - ETA: 5s - loss: 0.0107 - acc: 0.998 - ETA: 5s - loss: 0.0105 - acc: 0.998 - ETA: 5s - loss: 0.0104 - acc: 0.998 - ETA: 5s - loss: 0.0101 - acc: 0.998 - ETA: 5s - loss: 0.0095 - acc: 0.998 - ETA: 5s - loss: 0.0093 - acc: 0.998 - ETA: 5s - loss: 0.0091 - acc: 0.998 - ETA: 4s - loss: 0.0088 - acc: 0.998 - ETA: 4s - loss: 0.0089 - acc: 0.998 - ETA: 4s - loss: 0.0085 - acc: 0.998 - ETA: 4s - loss: 0.0082 - acc: 0.998 - ETA: 4s - loss: 0.0080 - acc: 0.998 - ETA: 4s - loss: 0.0096 - acc: 0.998 - ETA: 4s - loss: 0.0094 - acc: 0.998 - ETA: 4s - loss: 0.0092 - acc: 0.998 - ETA: 4s - loss: 0.0090 - acc: 0.998 - ETA: 4s - loss: 0.0089 - acc: 0.998 - ETA: 4s - loss: 0.0088 - acc: 0.998 - ETA: 4s - loss: 0.0086 - acc: 0.998 - ETA: 4s - loss: 0.0084 - acc: 0.998 - ETA: 3s - loss: 0.0086 - acc: 0.998 - ETA: 3s - loss: 0.0084 - acc: 0.998 - ETA: 3s - loss: 0.0082 - acc: 0.998 - ETA: 3s - loss: 0.0080 - acc: 0.998 - ETA: 3s - loss: 0.0079 - acc: 0.998 - ETA: 3s - loss: 0.0088 - acc: 0.997 - ETA: 3s - loss: 0.0099 - acc: 0.997 - ETA: 3s - loss: 0.0097 - acc: 0.997 - ETA: 3s - loss: 0.0095 - acc: 0.997 - ETA: 3s - loss: 0.0094 - acc: 0.997 - ETA: 3s - loss: 0.0092 - acc: 0.997 - ETA: 3s - loss: 0.0092 - acc: 0.997 - ETA: 3s - loss: 0.0091 - acc: 0.997 - ETA: 2s - loss: 0.0090 - acc: 0.997 - ETA: 2s - loss: 0.0089 - acc: 0.998 - ETA: 2s - loss: 0.0089 - acc: 0.998 - ETA: 2s - loss: 0.0088 - acc: 0.998 - ETA: 2s - loss: 0.0087 - acc: 0.998 - ETA: 2s - loss: 0.0086 - acc: 0.998 - ETA: 2s - loss: 0.0084 - acc: 0.998 - ETA: 2s - loss: 0.0099 - acc: 0.997 - ETA: 2s - loss: 0.0098 - acc: 0.997 - ETA: 2s - loss: 0.0097 - acc: 0.998 - ETA: 2s - loss: 0.0096 - acc: 0.998 - ETA: 2s - loss: 0.0095 - acc: 0.998 - ETA: 2s - loss: 0.0094 - acc: 0.998 - ETA: 2s - loss: 0.0095 - acc: 0.998 - ETA: 2s - loss: 0.0095 - acc: 0.998 - ETA: 2s - loss: 0.0094 - acc: 0.998 - ETA: 2s - loss: 0.0093 - acc: 0.998 - ETA: 2s - loss: 0.0092 - acc: 0.998 - ETA: 2s - loss: 0.0092 - acc: 0.998 - ETA: 1s - loss: 0.0092 - acc: 0.998 - ETA: 1s - loss: 0.0092 - acc: 0.998 - ETA: 1s - loss: 0.0103 - acc: 0.997 - ETA: 1s - loss: 0.0102 - acc: 0.997 - ETA: 1s - loss: 0.0101 - acc: 0.997 - ETA: 1s - loss: 0.0100 - acc: 0.998 - ETA: 1s - loss: 0.0099 - acc: 0.998 - ETA: 1s - loss: 0.0098 - acc: 0.998 - ETA: 1s - loss: 0.0097 - acc: 0.998 - ETA: 1s - loss: 0.0097 - acc: 0.998 - ETA: 1s - loss: 0.0099 - acc: 0.997 - ETA: 1s - loss: 0.0098 - acc: 0.997 - ETA: 1s - loss: 0.0097 - acc: 0.997 - ETA: 1s - loss: 0.0097 - acc: 0.997 - ETA: 1s - loss: 0.0097 - acc: 0.997 - ETA: 1s - loss: 0.0097 - acc: 0.997 - ETA: 1s - loss: 0.0096 - acc: 0.997 - ETA: 0s - loss: 0.0097 - acc: 0.997 - ETA: 0s - loss: 0.0096 - acc: 0.997 - ETA: 0s - loss: 0.0096 - acc: 0.997 - ETA: 0s - loss: 0.0096 - acc: 0.997 - ETA: 0s - loss: 0.0095 - acc: 0.997 - ETA: 0s - loss: 0.0098 - acc: 0.997 - ETA: 0s - loss: 0.0102 - acc: 0.997 - ETA: 0s - loss: 0.0102 - acc: 0.997 - ETA: 0s - loss: 0.0101 - acc: 0.997 - ETA: 0s - loss: 0.0100 - acc: 0.997 - ETA: 0s - loss: 0.0100 - acc: 0.997 - ETA: 0s - loss: 0.0100 - acc: 0.997 - ETA: 0s - loss: 0.0100 - acc: 0.997 - ETA: 0s - loss: 0.0100 - acc: 0.997 - ETA: 0s - loss: 0.0100 - acc: 0.997 - ETA: 0s - loss: 0.0100 - acc: 0.997 - ETA: 0s - loss: 0.0100 - acc: 0.997 - ETA: 0s - loss: 0.0102 - acc: 0.997 - ETA: 0s - loss: 0.0102 - acc: 0.997 - ETA: 0s - loss: 0.0101 - acc: 0.9977Epoch 00014: val_loss did not improve 6680/6680 [==============================] - 6s - loss: 0.0100 - acc: 0.9978 - val_loss: 0.8308 - val_acc: 0.8156 Epoch 16/20 6660/6680 [============================>.] - ETA: 6s - loss: 3.1661e-04 - acc: 1.000 - ETA: 5s - loss: 0.0028 - acc: 1.0000 - ETA: 5s - loss: 0.0022 - acc: 1.000 - ETA: 5s - loss: 0.0019 - acc: 1.000 - ETA: 5s - loss: 0.0019 - acc: 1.000 - ETA: 5s - loss: 0.0017 - acc: 1.000 - ETA: 5s - loss: 0.0016 - acc: 1.000 - ETA: 5s - loss: 0.0015 - acc: 1.000 - ETA: 5s - loss: 0.0014 - acc: 1.000 - ETA: 5s - loss: 0.0014 - acc: 1.000 - ETA: 5s - loss: 0.0014 - acc: 1.000 - ETA: 5s - loss: 0.0014 - acc: 1.000 - ETA: 5s - loss: 0.0015 - acc: 1.000 - ETA: 5s - loss: 0.0015 - acc: 1.000 - ETA: 5s - loss: 0.0046 - acc: 0.998 - ETA: 5s - loss: 0.0044 - acc: 0.998 - ETA: 5s - loss: 0.0042 - acc: 0.999 - ETA: 4s - loss: 0.0046 - acc: 0.999 - ETA: 4s - loss: 0.0045 - acc: 0.999 - ETA: 4s - loss: 0.0045 - acc: 0.999 - ETA: 4s - loss: 0.0044 - acc: 0.999 - ETA: 4s - loss: 0.0043 - acc: 0.999 - ETA: 4s - loss: 0.0042 - acc: 0.999 - ETA: 4s - loss: 0.0042 - acc: 0.999 - ETA: 4s - loss: 0.0046 - acc: 0.998 - ETA: 4s - loss: 0.0046 - acc: 0.998 - ETA: 4s - loss: 0.0045 - acc: 0.998 - ETA: 4s - loss: 0.0043 - acc: 0.998 - ETA: 4s - loss: 0.0042 - acc: 0.998 - ETA: 4s - loss: 0.0041 - acc: 0.998 - ETA: 4s - loss: 0.0040 - acc: 0.998 - ETA: 4s - loss: 0.0040 - acc: 0.999 - ETA: 4s - loss: 0.0040 - acc: 0.999 - ETA: 4s - loss: 0.0040 - acc: 0.999 - ETA: 4s - loss: 0.0057 - acc: 0.998 - ETA: 4s - loss: 0.0058 - acc: 0.998 - ETA: 4s - loss: 0.0057 - acc: 0.998 - ETA: 4s - loss: 0.0057 - acc: 0.998 - ETA: 4s - loss: 0.0067 - acc: 0.998 - ETA: 4s - loss: 0.0066 - acc: 0.998 - ETA: 3s - loss: 0.0065 - acc: 0.998 - ETA: 3s - loss: 0.0064 - acc: 0.998 - ETA: 3s - loss: 0.0067 - acc: 0.998 - ETA: 3s - loss: 0.0072 - acc: 0.997 - ETA: 3s - loss: 0.0071 - acc: 0.997 - ETA: 3s - loss: 0.0070 - acc: 0.997 - ETA: 3s - loss: 0.0070 - acc: 0.997 - ETA: 3s - loss: 0.0069 - acc: 0.997 - ETA: 3s - loss: 0.0069 - acc: 0.997 - ETA: 3s - loss: 0.0068 - acc: 0.997 - ETA: 3s - loss: 0.0068 - acc: 0.997 - ETA: 3s - loss: 0.0066 - acc: 0.998 - ETA: 3s - loss: 0.0066 - acc: 0.998 - ETA: 3s - loss: 0.0066 - acc: 0.998 - ETA: 3s - loss: 0.0066 - acc: 0.998 - ETA: 3s - loss: 0.0066 - acc: 0.998 - ETA: 3s - loss: 0.0066 - acc: 0.998 - ETA: 3s - loss: 0.0066 - acc: 0.998 - ETA: 3s - loss: 0.0065 - acc: 0.998 - ETA: 3s - loss: 0.0068 - acc: 0.997 - ETA: 3s - loss: 0.0067 - acc: 0.997 - ETA: 3s - loss: 0.0066 - acc: 0.997 - ETA: 3s - loss: 0.0065 - acc: 0.998 - ETA: 3s - loss: 0.0064 - acc: 0.998 - ETA: 3s - loss: 0.0063 - acc: 0.998 - ETA: 3s - loss: 0.0063 - acc: 0.998 - ETA: 3s - loss: 0.0062 - acc: 0.998 - ETA: 3s - loss: 0.0061 - acc: 0.998 - ETA: 2s - loss: 0.0060 - acc: 0.998 - ETA: 2s - loss: 0.0060 - acc: 0.998 - ETA: 2s - loss: 0.0060 - acc: 0.998 - ETA: 2s - loss: 0.0061 - acc: 0.998 - ETA: 2s - loss: 0.0060 - acc: 0.998 - ETA: 2s - loss: 0.0061 - acc: 0.998 - ETA: 2s - loss: 0.0060 - acc: 0.998 - ETA: 2s - loss: 0.0078 - acc: 0.998 - ETA: 2s - loss: 0.0077 - acc: 0.998 - ETA: 2s - loss: 0.0076 - acc: 0.998 - ETA: 2s - loss: 0.0075 - acc: 0.998 - ETA: 2s - loss: 0.0074 - acc: 0.998 - ETA: 2s - loss: 0.0074 - acc: 0.998 - ETA: 2s - loss: 0.0073 - acc: 0.998 - ETA: 2s - loss: 0.0072 - acc: 0.998 - ETA: 2s - loss: 0.0071 - acc: 0.998 - ETA: 1s - loss: 0.0075 - acc: 0.998 - ETA: 1s - loss: 0.0075 - acc: 0.998 - ETA: 1s - loss: 0.0075 - acc: 0.998 - ETA: 1s - loss: 0.0075 - acc: 0.998 - ETA: 1s - loss: 0.0074 - acc: 0.998 - ETA: 1s - loss: 0.0074 - acc: 0.998 - ETA: 1s - loss: 0.0074 - acc: 0.998 - ETA: 1s - loss: 0.0074 - acc: 0.998 - ETA: 1s - loss: 0.0080 - acc: 0.998 - ETA: 1s - loss: 0.0079 - acc: 0.998 - ETA: 1s - loss: 0.0078 - acc: 0.998 - ETA: 1s - loss: 0.0078 - acc: 0.998 - ETA: 1s - loss: 0.0078 - acc: 0.998 - ETA: 1s - loss: 0.0077 - acc: 0.998 - ETA: 1s - loss: 0.0077 - acc: 0.998 - ETA: 1s - loss: 0.0076 - acc: 0.998 - ETA: 1s - loss: 0.0075 - acc: 0.998 - ETA: 1s - loss: 0.0075 - acc: 0.998 - ETA: 0s - loss: 0.0074 - acc: 0.998 - ETA: 0s - loss: 0.0074 - acc: 0.998 - ETA: 0s - loss: 0.0073 - acc: 0.998 - ETA: 0s - loss: 0.0073 - acc: 0.998 - ETA: 0s - loss: 0.0073 - acc: 0.998 - ETA: 0s - loss: 0.0072 - acc: 0.998 - ETA: 0s - loss: 0.0072 - acc: 0.998 - ETA: 0s - loss: 0.0072 - acc: 0.998 - ETA: 0s - loss: 0.0071 - acc: 0.998 - ETA: 0s - loss: 0.0071 - acc: 0.998 - ETA: 0s - loss: 0.0071 - acc: 0.998 - ETA: 0s - loss: 0.0071 - acc: 0.998 - ETA: 0s - loss: 0.0071 - acc: 0.998 - ETA: 0s - loss: 0.0079 - acc: 0.998 - ETA: 0s - loss: 0.0079 - acc: 0.998 - ETA: 0s - loss: 0.0081 - acc: 0.998 - ETA: 0s - loss: 0.0081 - acc: 0.998 - ETA: 0s - loss: 0.0080 - acc: 0.998 - ETA: 0s - loss: 0.0080 - acc: 0.998 - ETA: 0s - loss: 0.0080 - acc: 0.998 - ETA: 0s - loss: 0.0080 - acc: 0.9982Epoch 00015: val_loss did not improve 6680/6680 [==============================] - 7s - loss: 0.0080 - acc: 0.9982 - val_loss: 0.9199 - val_acc: 0.8204 Epoch 17/20 6640/6680 [============================>.] - ETA: 6s - loss: 0.0016 - acc: 1.000 - ETA: 5s - loss: 9.1708e-04 - acc: 1.000 - ETA: 5s - loss: 0.0011 - acc: 1.0000 - ETA: 5s - loss: 8.9403e-04 - acc: 1.000 - ETA: 5s - loss: 0.0039 - acc: 0.9964 - ETA: 6s - loss: 0.0040 - acc: 0.996 - ETA: 6s - loss: 0.0034 - acc: 0.997 - ETA: 6s - loss: 0.0030 - acc: 0.997 - ETA: 5s - loss: 0.0039 - acc: 0.998 - ETA: 5s - loss: 0.0035 - acc: 0.998 - ETA: 5s - loss: 0.0035 - acc: 0.998 - ETA: 5s - loss: 0.0033 - acc: 0.998 - ETA: 5s - loss: 0.0031 - acc: 0.998 - ETA: 5s - loss: 0.0029 - acc: 0.998 - ETA: 5s - loss: 0.0031 - acc: 0.998 - ETA: 5s - loss: 0.0030 - acc: 0.998 - ETA: 5s - loss: 0.0029 - acc: 0.999 - ETA: 5s - loss: 0.0028 - acc: 0.999 - ETA: 5s - loss: 0.0027 - acc: 0.999 - ETA: 5s - loss: 0.0026 - acc: 0.999 - ETA: 4s - loss: 0.0025 - acc: 0.999 - ETA: 4s - loss: 0.0024 - acc: 0.999 - ETA: 4s - loss: 0.0025 - acc: 0.999 - ETA: 4s - loss: 0.0025 - acc: 0.999 - ETA: 4s - loss: 0.0024 - acc: 0.999 - ETA: 4s - loss: 0.0024 - acc: 0.999 - ETA: 4s - loss: 0.0024 - acc: 0.999 - ETA: 4s - loss: 0.0023 - acc: 0.999 - ETA: 4s - loss: 0.0023 - acc: 0.999 - ETA: 4s - loss: 0.0024 - acc: 0.999 - ETA: 4s - loss: 0.0023 - acc: 0.999 - ETA: 4s - loss: 0.0025 - acc: 0.999 - ETA: 4s - loss: 0.0025 - acc: 0.999 - ETA: 4s - loss: 0.0025 - acc: 0.999 - ETA: 4s - loss: 0.0025 - acc: 0.999 - ETA: 4s - loss: 0.0024 - acc: 0.999 - ETA: 4s - loss: 0.0024 - acc: 0.999 - ETA: 4s - loss: 0.0023 - acc: 0.999 - ETA: 3s - loss: 0.0023 - acc: 0.999 - ETA: 3s - loss: 0.0022 - acc: 0.999 - ETA: 3s - loss: 0.0022 - acc: 0.999 - ETA: 3s - loss: 0.0023 - acc: 0.999 - ETA: 3s - loss: 0.0023 - acc: 0.999 - ETA: 3s - loss: 0.0033 - acc: 0.999 - ETA: 3s - loss: 0.0033 - acc: 0.999 - ETA: 3s - loss: 0.0033 - acc: 0.999 - ETA: 3s - loss: 0.0033 - acc: 0.999 - ETA: 3s - loss: 0.0033 - acc: 0.999 - ETA: 3s - loss: 0.0032 - acc: 0.999 - ETA: 3s - loss: 0.0032 - acc: 0.999 - ETA: 3s - loss: 0.0034 - acc: 0.999 - ETA: 3s - loss: 0.0033 - acc: 0.999 - ETA: 3s - loss: 0.0033 - acc: 0.999 - ETA: 3s - loss: 0.0033 - acc: 0.999 - ETA: 3s - loss: 0.0033 - acc: 0.999 - ETA: 3s - loss: 0.0034 - acc: 0.999 - ETA: 3s - loss: 0.0038 - acc: 0.999 - ETA: 3s - loss: 0.0037 - acc: 0.999 - ETA: 3s - loss: 0.0037 - acc: 0.999 - ETA: 3s - loss: 0.0036 - acc: 0.999 - ETA: 3s - loss: 0.0037 - acc: 0.999 - ETA: 2s - loss: 0.0037 - acc: 0.999 - ETA: 2s - loss: 0.0036 - acc: 0.999 - ETA: 2s - loss: 0.0036 - acc: 0.999 - ETA: 2s - loss: 0.0035 - acc: 0.999 - ETA: 2s - loss: 0.0035 - acc: 0.999 - ETA: 2s - loss: 0.0035 - acc: 0.999 - ETA: 2s - loss: 0.0035 - acc: 0.999 - ETA: 2s - loss: 0.0035 - acc: 0.999 - ETA: 2s - loss: 0.0035 - acc: 0.999 - ETA: 2s - loss: 0.0034 - acc: 0.999 - ETA: 2s - loss: 0.0034 - acc: 0.999 - ETA: 2s - loss: 0.0034 - acc: 0.999 - ETA: 2s - loss: 0.0035 - acc: 0.999 - ETA: 2s - loss: 0.0047 - acc: 0.999 - ETA: 2s - loss: 0.0046 - acc: 0.999 - ETA: 1s - loss: 0.0049 - acc: 0.998 - ETA: 1s - loss: 0.0048 - acc: 0.998 - ETA: 1s - loss: 0.0057 - acc: 0.998 - ETA: 1s - loss: 0.0056 - acc: 0.998 - ETA: 1s - loss: 0.0055 - acc: 0.998 - ETA: 1s - loss: 0.0055 - acc: 0.998 - ETA: 1s - loss: 0.0055 - acc: 0.998 - ETA: 1s - loss: 0.0054 - acc: 0.998 - ETA: 1s - loss: 0.0054 - acc: 0.998 - ETA: 1s - loss: 0.0054 - acc: 0.998 - ETA: 1s - loss: 0.0053 - acc: 0.998 - ETA: 1s - loss: 0.0057 - acc: 0.998 - ETA: 1s - loss: 0.0057 - acc: 0.998 - ETA: 1s - loss: 0.0057 - acc: 0.998 - ETA: 1s - loss: 0.0056 - acc: 0.998 - ETA: 1s - loss: 0.0060 - acc: 0.998 - ETA: 1s - loss: 0.0060 - acc: 0.998 - ETA: 0s - loss: 0.0059 - acc: 0.998 - ETA: 0s - loss: 0.0059 - acc: 0.998 - ETA: 0s - loss: 0.0063 - acc: 0.998 - ETA: 0s - loss: 0.0062 - acc: 0.998 - ETA: 0s - loss: 0.0062 - acc: 0.998 - ETA: 0s - loss: 0.0061 - acc: 0.998 - ETA: 0s - loss: 0.0061 - acc: 0.998 - ETA: 0s - loss: 0.0061 - acc: 0.998 - ETA: 0s - loss: 0.0061 - acc: 0.998 - ETA: 0s - loss: 0.0061 - acc: 0.998 - ETA: 0s - loss: 0.0060 - acc: 0.998 - ETA: 0s - loss: 0.0061 - acc: 0.998 - ETA: 0s - loss: 0.0060 - acc: 0.998 - ETA: 0s - loss: 0.0060 - acc: 0.998 - ETA: 0s - loss: 0.0059 - acc: 0.998 - ETA: 0s - loss: 0.0059 - acc: 0.998 - ETA: 0s - loss: 0.0071 - acc: 0.998 - ETA: 0s - loss: 0.0070 - acc: 0.998 - ETA: 0s - loss: 0.0070 - acc: 0.998 - ETA: 0s - loss: 0.0072 - acc: 0.9982Epoch 00016: val_loss did not improve 6680/6680 [==============================] - 6s - loss: 0.0072 - acc: 0.9982 - val_loss: 0.8625 - val_acc: 0.8275 Epoch 18/20 6640/6680 [============================>.] - ETA: 6s - loss: 1.4823e-04 - acc: 1.000 - ETA: 5s - loss: 0.0032 - acc: 1.0000 - ETA: 5s - loss: 0.0075 - acc: 1.000 - ETA: 5s - loss: 0.0055 - acc: 1.000 - ETA: 5s - loss: 0.0044 - acc: 1.000 - ETA: 5s - loss: 0.0036 - acc: 1.000 - ETA: 5s - loss: 0.0032 - acc: 1.000 - ETA: 5s - loss: 0.0031 - acc: 1.000 - ETA: 5s - loss: 0.0028 - acc: 1.000 - ETA: 5s - loss: 0.0026 - acc: 1.000 - ETA: 5s - loss: 0.0024 - acc: 1.000 - ETA: 5s - loss: 0.0023 - acc: 1.000 - ETA: 5s - loss: 0.0021 - acc: 1.000 - ETA: 5s - loss: 0.0021 - acc: 1.000 - ETA: 6s - loss: 0.0020 - acc: 1.000 - ETA: 6s - loss: 0.0019 - acc: 1.000 - ETA: 6s - loss: 0.0018 - acc: 1.000 - ETA: 5s - loss: 0.0018 - acc: 1.000 - ETA: 5s - loss: 0.0042 - acc: 0.999 - ETA: 5s - loss: 0.0042 - acc: 0.999 - ETA: 5s - loss: 0.0040 - acc: 0.999 - ETA: 5s - loss: 0.0038 - acc: 0.999 - ETA: 5s - loss: 0.0036 - acc: 0.999 - ETA: 5s - loss: 0.0035 - acc: 0.999 - ETA: 5s - loss: 0.0034 - acc: 0.999 - ETA: 5s - loss: 0.0032 - acc: 0.999 - ETA: 5s - loss: 0.0031 - acc: 0.999 - ETA: 5s - loss: 0.0030 - acc: 0.999 - ETA: 4s - loss: 0.0030 - acc: 0.999 - ETA: 4s - loss: 0.0030 - acc: 0.999 - ETA: 4s - loss: 0.0029 - acc: 0.999 - ETA: 4s - loss: 0.0028 - acc: 0.999 - ETA: 4s - loss: 0.0028 - acc: 0.999 - ETA: 4s - loss: 0.0027 - acc: 0.999 - ETA: 4s - loss: 0.0029 - acc: 0.999 - ETA: 4s - loss: 0.0060 - acc: 0.999 - ETA: 4s - loss: 0.0058 - acc: 0.999 - ETA: 4s - loss: 0.0057 - acc: 0.999 - ETA: 4s - loss: 0.0057 - acc: 0.999 - ETA: 5s - loss: 0.0056 - acc: 0.999 - ETA: 5s - loss: 0.0056 - acc: 0.999 - ETA: 4s - loss: 0.0054 - acc: 0.999 - ETA: 4s - loss: 0.0068 - acc: 0.998 - ETA: 4s - loss: 0.0066 - acc: 0.998 - ETA: 4s - loss: 0.0065 - acc: 0.998 - ETA: 4s - loss: 0.0064 - acc: 0.998 - ETA: 4s - loss: 0.0063 - acc: 0.998 - ETA: 4s - loss: 0.0062 - acc: 0.998 - ETA: 4s - loss: 0.0061 - acc: 0.998 - ETA: 4s - loss: 0.0059 - acc: 0.998 - ETA: 4s - loss: 0.0058 - acc: 0.998 - ETA: 4s - loss: 0.0057 - acc: 0.998 - ETA: 4s - loss: 0.0056 - acc: 0.998 - ETA: 3s - loss: 0.0055 - acc: 0.999 - ETA: 3s - loss: 0.0054 - acc: 0.999 - ETA: 3s - loss: 0.0053 - acc: 0.999 - ETA: 3s - loss: 0.0052 - acc: 0.999 - ETA: 3s - loss: 0.0052 - acc: 0.999 - ETA: 3s - loss: 0.0051 - acc: 0.999 - ETA: 3s - loss: 0.0052 - acc: 0.999 - ETA: 3s - loss: 0.0051 - acc: 0.999 - ETA: 3s - loss: 0.0050 - acc: 0.999 - ETA: 3s - loss: 0.0050 - acc: 0.999 - ETA: 3s - loss: 0.0049 - acc: 0.999 - ETA: 3s - loss: 0.0048 - acc: 0.999 - ETA: 3s - loss: 0.0048 - acc: 0.999 - ETA: 2s - loss: 0.0047 - acc: 0.999 - ETA: 2s - loss: 0.0047 - acc: 0.999 - ETA: 2s - loss: 0.0046 - acc: 0.999 - ETA: 2s - loss: 0.0046 - acc: 0.999 - ETA: 2s - loss: 0.0045 - acc: 0.999 - ETA: 2s - loss: 0.0045 - acc: 0.999 - ETA: 2s - loss: 0.0045 - acc: 0.999 - ETA: 2s - loss: 0.0045 - acc: 0.999 - ETA: 2s - loss: 0.0044 - acc: 0.999 - ETA: 2s - loss: 0.0044 - acc: 0.999 - ETA: 2s - loss: 0.0044 - acc: 0.999 - ETA: 2s - loss: 0.0045 - acc: 0.999 - ETA: 2s - loss: 0.0045 - acc: 0.999 - ETA: 2s - loss: 0.0045 - acc: 0.999 - ETA: 2s - loss: 0.0044 - acc: 0.999 - ETA: 2s - loss: 0.0044 - acc: 0.999 - ETA: 2s - loss: 0.0044 - acc: 0.999 - ETA: 2s - loss: 0.0043 - acc: 0.999 - ETA: 2s - loss: 0.0043 - acc: 0.999 - ETA: 2s - loss: 0.0043 - acc: 0.999 - ETA: 2s - loss: 0.0043 - acc: 0.999 - ETA: 1s - loss: 0.0047 - acc: 0.999 - ETA: 1s - loss: 0.0047 - acc: 0.999 - ETA: 1s - loss: 0.0046 - acc: 0.999 - ETA: 1s - loss: 0.0052 - acc: 0.999 - ETA: 1s - loss: 0.0052 - acc: 0.999 - ETA: 1s - loss: 0.0070 - acc: 0.998 - ETA: 1s - loss: 0.0069 - acc: 0.998 - ETA: 1s - loss: 0.0068 - acc: 0.998 - ETA: 1s - loss: 0.0067 - acc: 0.998 - ETA: 1s - loss: 0.0066 - acc: 0.998 - ETA: 1s - loss: 0.0066 - acc: 0.998 - ETA: 1s - loss: 0.0065 - acc: 0.998 - ETA: 1s - loss: 0.0066 - acc: 0.998 - ETA: 1s - loss: 0.0066 - acc: 0.998 - ETA: 1s - loss: 0.0065 - acc: 0.998 - ETA: 0s - loss: 0.0065 - acc: 0.998 - ETA: 0s - loss: 0.0065 - acc: 0.999 - ETA: 0s - loss: 0.0064 - acc: 0.999 - ETA: 0s - loss: 0.0066 - acc: 0.998 - ETA: 0s - loss: 0.0066 - acc: 0.998 - ETA: 0s - loss: 0.0065 - acc: 0.998 - ETA: 0s - loss: 0.0065 - acc: 0.998 - ETA: 0s - loss: 0.0065 - acc: 0.998 - ETA: 0s - loss: 0.0072 - acc: 0.998 - ETA: 0s - loss: 0.0071 - acc: 0.998 - ETA: 0s - loss: 0.0071 - acc: 0.998 - ETA: 0s - loss: 0.0070 - acc: 0.998 - ETA: 0s - loss: 0.0070 - acc: 0.998 - ETA: 0s - loss: 0.0069 - acc: 0.998 - ETA: 0s - loss: 0.0081 - acc: 0.998 - ETA: 0s - loss: 0.0080 - acc: 0.998 - ETA: 0s - loss: 0.0080 - acc: 0.998 - ETA: 0s - loss: 0.0080 - acc: 0.998 - ETA: 0s - loss: 0.0086 - acc: 0.9983Epoch 00017: val_loss did not improve 6680/6680 [==============================] - 7s - loss: 0.0086 - acc: 0.9984 - val_loss: 0.9336 - val_acc: 0.8192 Epoch 19/20 6660/6680 [============================>.] - ETA: 7s - loss: 0.0013 - acc: 1.000 - ETA: 5s - loss: 5.3899e-04 - acc: 1.000 - ETA: 5s - loss: 8.3471e-04 - acc: 1.000 - ETA: 5s - loss: 0.0020 - acc: 1.0000 - ETA: 5s - loss: 0.0018 - acc: 1.000 - ETA: 5s - loss: 0.0015 - acc: 1.000 - ETA: 5s - loss: 0.0015 - acc: 1.000 - ETA: 5s - loss: 0.0176 - acc: 0.995 - ETA: 5s - loss: 0.0157 - acc: 0.996 - ETA: 5s - loss: 0.0142 - acc: 0.996 - ETA: 5s - loss: 0.0130 - acc: 0.997 - ETA: 5s - loss: 0.0120 - acc: 0.997 - ETA: 5s - loss: 0.0112 - acc: 0.997 - ETA: 5s - loss: 0.0106 - acc: 0.997 - ETA: 5s - loss: 0.0102 - acc: 0.997 - ETA: 5s - loss: 0.0097 - acc: 0.997 - ETA: 5s - loss: 0.0092 - acc: 0.997 - ETA: 5s - loss: 0.0087 - acc: 0.998 - ETA: 5s - loss: 0.0081 - acc: 0.998 - ETA: 5s - loss: 0.0077 - acc: 0.998 - ETA: 5s - loss: 0.0073 - acc: 0.998 - ETA: 4s - loss: 0.0070 - acc: 0.998 - ETA: 4s - loss: 0.0077 - acc: 0.997 - ETA: 4s - loss: 0.0075 - acc: 0.997 - ETA: 4s - loss: 0.0072 - acc: 0.998 - ETA: 4s - loss: 0.0069 - acc: 0.998 - ETA: 4s - loss: 0.0067 - acc: 0.998 - ETA: 4s - loss: 0.0065 - acc: 0.998 - ETA: 4s - loss: 0.0062 - acc: 0.998 - ETA: 4s - loss: 0.0061 - acc: 0.998 - ETA: 4s - loss: 0.0060 - acc: 0.998 - ETA: 4s - loss: 0.0059 - acc: 0.998 - ETA: 4s - loss: 0.0057 - acc: 0.998 - ETA: 4s - loss: 0.0056 - acc: 0.998 - ETA: 4s - loss: 0.0055 - acc: 0.998 - ETA: 4s - loss: 0.0055 - acc: 0.998 - ETA: 4s - loss: 0.0054 - acc: 0.998 - ETA: 4s - loss: 0.0053 - acc: 0.998 - ETA: 4s - loss: 0.0051 - acc: 0.998 - ETA: 4s - loss: 0.0050 - acc: 0.998 - ETA: 4s - loss: 0.0049 - acc: 0.998 - ETA: 4s - loss: 0.0048 - acc: 0.998 - ETA: 3s - loss: 0.0047 - acc: 0.998 - ETA: 3s - loss: 0.0046 - acc: 0.998 - ETA: 3s - loss: 0.0045 - acc: 0.998 - ETA: 3s - loss: 0.0045 - acc: 0.998 - ETA: 3s - loss: 0.0047 - acc: 0.998 - ETA: 3s - loss: 0.0046 - acc: 0.998 - ETA: 3s - loss: 0.0045 - acc: 0.999 - ETA: 3s - loss: 0.0044 - acc: 0.999 - ETA: 3s - loss: 0.0043 - acc: 0.999 - ETA: 3s - loss: 0.0042 - acc: 0.999 - ETA: 3s - loss: 0.0042 - acc: 0.999 - ETA: 3s - loss: 0.0041 - acc: 0.999 - ETA: 3s - loss: 0.0040 - acc: 0.999 - ETA: 3s - loss: 0.0040 - acc: 0.999 - ETA: 3s - loss: 0.0039 - acc: 0.999 - ETA: 2s - loss: 0.0039 - acc: 0.999 - ETA: 2s - loss: 0.0046 - acc: 0.998 - ETA: 2s - loss: 0.0046 - acc: 0.998 - ETA: 2s - loss: 0.0045 - acc: 0.998 - ETA: 2s - loss: 0.0044 - acc: 0.998 - ETA: 2s - loss: 0.0044 - acc: 0.999 - ETA: 2s - loss: 0.0044 - acc: 0.999 - ETA: 2s - loss: 0.0044 - acc: 0.999 - ETA: 2s - loss: 0.0043 - acc: 0.999 - ETA: 2s - loss: 0.0043 - acc: 0.999 - ETA: 2s - loss: 0.0042 - acc: 0.999 - ETA: 2s - loss: 0.0042 - acc: 0.999 - ETA: 2s - loss: 0.0042 - acc: 0.999 - ETA: 2s - loss: 0.0041 - acc: 0.999 - ETA: 2s - loss: 0.0041 - acc: 0.999 - ETA: 2s - loss: 0.0043 - acc: 0.998 - ETA: 2s - loss: 0.0045 - acc: 0.998 - ETA: 1s - loss: 0.0056 - acc: 0.998 - ETA: 1s - loss: 0.0061 - acc: 0.998 - ETA: 1s - loss: 0.0060 - acc: 0.998 - ETA: 1s - loss: 0.0060 - acc: 0.998 - ETA: 1s - loss: 0.0059 - acc: 0.998 - ETA: 1s - loss: 0.0059 - acc: 0.998 - ETA: 1s - loss: 0.0058 - acc: 0.998 - ETA: 1s - loss: 0.0057 - acc: 0.998 - ETA: 1s - loss: 0.0056 - acc: 0.998 - ETA: 1s - loss: 0.0056 - acc: 0.998 - ETA: 1s - loss: 0.0056 - acc: 0.998 - ETA: 1s - loss: 0.0055 - acc: 0.998 - ETA: 1s - loss: 0.0055 - acc: 0.998 - ETA: 1s - loss: 0.0055 - acc: 0.998 - ETA: 1s - loss: 0.0054 - acc: 0.998 - ETA: 1s - loss: 0.0054 - acc: 0.998 - ETA: 1s - loss: 0.0053 - acc: 0.998 - ETA: 1s - loss: 0.0053 - acc: 0.998 - ETA: 0s - loss: 0.0053 - acc: 0.998 - ETA: 0s - loss: 0.0053 - acc: 0.998 - ETA: 0s - loss: 0.0053 - acc: 0.998 - ETA: 0s - loss: 0.0052 - acc: 0.998 - ETA: 0s - loss: 0.0055 - acc: 0.998 - ETA: 0s - loss: 0.0054 - acc: 0.998 - ETA: 0s - loss: 0.0055 - acc: 0.998 - ETA: 0s - loss: 0.0055 - acc: 0.998 - ETA: 0s - loss: 0.0055 - acc: 0.998 - ETA: 0s - loss: 0.0054 - acc: 0.998 - ETA: 0s - loss: 0.0054 - acc: 0.998 - ETA: 0s - loss: 0.0054 - acc: 0.998 - ETA: 0s - loss: 0.0054 - acc: 0.998 - ETA: 0s - loss: 0.0054 - acc: 0.998 - ETA: 0s - loss: 0.0054 - acc: 0.998 - ETA: 0s - loss: 0.0053 - acc: 0.998 - ETA: 0s - loss: 0.0053 - acc: 0.998 - ETA: 0s - loss: 0.0054 - acc: 0.9985Epoch 00018: val_loss did not improve 6680/6680 [==============================] - 6s - loss: 0.0054 - acc: 0.9985 - val_loss: 0.8973 - val_acc: 0.8287 Epoch 20/20 6640/6680 [============================>.] - ETA: 6s - loss: 6.1588e-05 - acc: 1.000 - ETA: 6s - loss: 9.3723e-04 - acc: 1.000 - ETA: 6s - loss: 5.9945e-04 - acc: 1.000 - ETA: 5s - loss: 4.8767e-04 - acc: 1.000 - ETA: 5s - loss: 0.0093 - acc: 0.9962 - ETA: 5s - loss: 0.0079 - acc: 0.996 - ETA: 5s - loss: 0.0067 - acc: 0.997 - ETA: 6s - loss: 0.0159 - acc: 0.995 - ETA: 6s - loss: 0.0145 - acc: 0.995 - ETA: 6s - loss: 0.0229 - acc: 0.994 - ETA: 6s - loss: 0.0213 - acc: 0.994 - ETA: 6s - loss: 0.0207 - acc: 0.994 - ETA: 6s - loss: 0.0187 - acc: 0.995 - ETA: 6s - loss: 0.0172 - acc: 0.995 - ETA: 6s - loss: 0.0158 - acc: 0.995 - ETA: 6s - loss: 0.0146 - acc: 0.996 - ETA: 6s - loss: 0.0136 - acc: 0.996 - ETA: 5s - loss: 0.0128 - acc: 0.996 - ETA: 5s - loss: 0.0121 - acc: 0.997 - ETA: 5s - loss: 0.0114 - acc: 0.997 - ETA: 5s - loss: 0.0109 - acc: 0.997 - ETA: 5s - loss: 0.0104 - acc: 0.997 - ETA: 5s - loss: 0.0099 - acc: 0.997 - ETA: 5s - loss: 0.0095 - acc: 0.997 - ETA: 5s - loss: 0.0091 - acc: 0.997 - ETA: 5s - loss: 0.0087 - acc: 0.997 - ETA: 5s - loss: 0.0085 - acc: 0.998 - ETA: 5s - loss: 0.0082 - acc: 0.998 - ETA: 4s - loss: 0.0079 - acc: 0.998 - ETA: 4s - loss: 0.0076 - acc: 0.998 - ETA: 4s - loss: 0.0073 - acc: 0.998 - ETA: 4s - loss: 0.0071 - acc: 0.998 - ETA: 4s - loss: 0.0083 - acc: 0.997 - ETA: 4s - loss: 0.0080 - acc: 0.997 - ETA: 4s - loss: 0.0077 - acc: 0.998 - ETA: 4s - loss: 0.0075 - acc: 0.998 - ETA: 4s - loss: 0.0072 - acc: 0.998 - ETA: 4s - loss: 0.0071 - acc: 0.998 - ETA: 4s - loss: 0.0070 - acc: 0.998 - ETA: 4s - loss: 0.0069 - acc: 0.998 - ETA: 4s - loss: 0.0068 - acc: 0.998 - ETA: 3s - loss: 0.0067 - acc: 0.998 - ETA: 3s - loss: 0.0073 - acc: 0.998 - ETA: 3s - loss: 0.0072 - acc: 0.998 - ETA: 3s - loss: 0.0070 - acc: 0.998 - ETA: 3s - loss: 0.0069 - acc: 0.998 - ETA: 3s - loss: 0.0068 - acc: 0.998 - ETA: 3s - loss: 0.0067 - acc: 0.998 - ETA: 3s - loss: 0.0065 - acc: 0.998 - ETA: 3s - loss: 0.0064 - acc: 0.998 - ETA: 3s - loss: 0.0063 - acc: 0.998 - ETA: 3s - loss: 0.0062 - acc: 0.998 - ETA: 3s - loss: 0.0061 - acc: 0.998 - ETA: 3s - loss: 0.0060 - acc: 0.998 - ETA: 3s - loss: 0.0059 - acc: 0.998 - ETA: 3s - loss: 0.0063 - acc: 0.998 - ETA: 3s - loss: 0.0062 - acc: 0.998 - ETA: 3s - loss: 0.0061 - acc: 0.998 - ETA: 2s - loss: 0.0060 - acc: 0.998 - ETA: 2s - loss: 0.0058 - acc: 0.998 - ETA: 2s - loss: 0.0058 - acc: 0.998 - ETA: 2s - loss: 0.0057 - acc: 0.998 - ETA: 2s - loss: 0.0056 - acc: 0.998 - ETA: 2s - loss: 0.0054 - acc: 0.998 - ETA: 2s - loss: 0.0054 - acc: 0.998 - ETA: 2s - loss: 0.0053 - acc: 0.998 - ETA: 2s - loss: 0.0052 - acc: 0.998 - ETA: 2s - loss: 0.0051 - acc: 0.998 - ETA: 2s - loss: 0.0050 - acc: 0.998 - ETA: 2s - loss: 0.0050 - acc: 0.998 - ETA: 2s - loss: 0.0049 - acc: 0.998 - ETA: 1s - loss: 0.0048 - acc: 0.998 - ETA: 1s - loss: 0.0051 - acc: 0.998 - ETA: 1s - loss: 0.0050 - acc: 0.998 - ETA: 1s - loss: 0.0050 - acc: 0.998 - ETA: 1s - loss: 0.0053 - acc: 0.998 - ETA: 1s - loss: 0.0052 - acc: 0.998 - ETA: 1s - loss: 0.0052 - acc: 0.998 - ETA: 1s - loss: 0.0052 - acc: 0.998 - ETA: 1s - loss: 0.0051 - acc: 0.998 - ETA: 1s - loss: 0.0051 - acc: 0.998 - ETA: 1s - loss: 0.0051 - acc: 0.998 - ETA: 1s - loss: 0.0050 - acc: 0.998 - ETA: 1s - loss: 0.0050 - acc: 0.998 - ETA: 1s - loss: 0.0056 - acc: 0.998 - ETA: 1s - loss: 0.0056 - acc: 0.998 - ETA: 1s - loss: 0.0055 - acc: 0.998 - ETA: 1s - loss: 0.0054 - acc: 0.998 - ETA: 1s - loss: 0.0054 - acc: 0.998 - ETA: 1s - loss: 0.0053 - acc: 0.998 - ETA: 1s - loss: 0.0053 - acc: 0.998 - ETA: 0s - loss: 0.0052 - acc: 0.998 - ETA: 0s - loss: 0.0052 - acc: 0.998 - ETA: 0s - loss: 0.0052 - acc: 0.998 - ETA: 0s - loss: 0.0051 - acc: 0.998 - ETA: 0s - loss: 0.0051 - acc: 0.998 - ETA: 0s - loss: 0.0051 - acc: 0.998 - ETA: 0s - loss: 0.0050 - acc: 0.998 - ETA: 0s - loss: 0.0050 - acc: 0.998 - ETA: 0s - loss: 0.0050 - acc: 0.998 - ETA: 0s - loss: 0.0049 - acc: 0.998 - ETA: 0s - loss: 0.0049 - acc: 0.998 - ETA: 0s - loss: 0.0050 - acc: 0.998 - ETA: 0s - loss: 0.0050 - acc: 0.998 - ETA: 0s - loss: 0.0049 - acc: 0.998 - ETA: 0s - loss: 0.0049 - acc: 0.998 - ETA: 0s - loss: 0.0048 - acc: 0.998 - ETA: 0s - loss: 0.0048 - acc: 0.998 - ETA: 0s - loss: 0.0048 - acc: 0.998 - ETA: 0s - loss: 0.0048 - acc: 0.998 - ETA: 0s - loss: 0.0048 - acc: 0.9985Epoch 00019: val_loss did not improve 6680/6680 [==============================] - 6s - loss: 0.0047 - acc: 0.9985 - val_loss: 0.9380 - val_acc: 0.8263
Out[28]:
<keras.callbacks.History at 0x23329802a20>
(IMPLEMENTATION) Load the Model with the Best Validation Loss¶
In [29]:
### TODO: Load the model weights with the best validation loss.
Resnet50_model.load_weights('saved_models/weights.best.ResNet50.hdf5')
(IMPLEMENTATION) Test the Model¶
Try out your model on the test dataset of dog images. Ensure that your test accuracy is greater than 60%.
In [30]:
### TODO: Calculate classification accuracy on the test dataset.
# get index of predicted dog breed for each image in test set
ResNet50_predictions = [np.argmax(Resnet50_model.predict(np.expand_dims(feature, axis=0))) for feature in test_ResNet50]
# report test accuracy
test_accuracy = 100*np.sum(np.array(ResNet50_predictions)==np.argmax(test_targets, axis=1))/len(ResNet50_predictions)
print('Test accuracy: %.4f%%' % test_accuracy)
Test accuracy: 80.5024%
(IMPLEMENTATION) Predict Dog Breed with the Model¶
Write a function that takes an image path as input and returns the dog breed (Affenpinscher, Afghan_hound, etc) that is predicted by your model.
Similar to the analogous function in Step 5, your function should have three steps:
- Extract the bottleneck features corresponding to the chosen CNN model.
- Supply the bottleneck features as input to the model to return the predicted vector. Note that the argmax of this prediction vector gives the index of the predicted dog breed.
- Use the
dog_namesarray defined in Step 0 of this notebook to return the corresponding breed.
The functions to extract the bottleneck features can be found in extract_bottleneck_features.py, and they have been imported in an earlier code cell. To obtain the bottleneck features corresponding to your chosen CNN architecture, you need to use the function
extract_{network}
where {network}, in the above filename, should be one of VGG19, Resnet50, InceptionV3, or Xception.
In [31]:
### TODO: Write a function that takes a path to an image as input
### and returns the dog breed that is predicted by the model.
from extract_bottleneck_features import *
def ResNet50_predict_breed(img_path):
# extract bottleneck features
bottleneck_feature = extract_Resnet50(path_to_tensor(img_path))
# obtain predicted vector
predicted_vector = Resnet50_model.predict(bottleneck_feature)
# return dog breed that is predicted by the model
return dog_names[np.argmax(predicted_vector)]
Step 6: Write your Algorithm¶
Write an algorithm that accepts a file path to an image and first determines whether the image contains a human, dog, or neither. Then,
- if a dog is detected in the image, return the predicted breed.
- if a human is detected in the image, return the resembling dog breed.
- if neither is detected in the image, provide output that indicates an error.
You are welcome to write your own functions for detecting humans and dogs in images, but feel free to use the face_detector and dog_detector functions developed above. You are required to use your CNN from Step 5 to predict dog breed.
Some sample output for our algorithm is provided below, but feel free to design your own user experience!
(IMPLEMENTATION) Write your Algorithm¶
In [32]:
#Algorithm
import numpy as np
from glob import glob
def dog_breed_detector(img_path):
breed = ResNet50_predict_breed(img_path)
# Display the image
img = cv2.imread(img_path)
cv_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
plt.imshow(cv_rgb)
plt.show()
# Detect what it is
if dog_detector(img_path):
print("That's a dog. Breed: " + str(breed))
elif face_detector(img_path):
print("That's a human, but it looks like a " + str(breed))
else:
print("Can't detect anything!")
## Load the cell
sample_files = np.array(glob("Test_image/*"))
print(sample_files)
['Test_image\\download (1).jpg' 'Test_image\\download.jpg' 'Test_image\\photo_2018-02-27_21-04-25.jpg' 'Test_image\\photo_2018-02-27_21-04-37.jpg' 'Test_image\\photo_2018-02-27_21-04-47.jpg' 'Test_image\\photo_2018-02-27_21-04-55.jpg' 'Test_image\\photo_2018-02-27_21-05-03.jpg' 'Test_image\\photo_2018-02-27_21-05-13.jpg' 'Test_image\\photo_2018-02-27_21-05-33.jpg']
Step 7: Test Your Algorithm¶
In this section, you will take your new algorithm for a spin! What kind of dog does the algorithm think that you look like? If you have a dog, does it predict your dog's breed accurately? If you have a cat, does it mistakenly think that your cat is a dog?
(IMPLEMENTATION) Test Your Algorithm on Sample Images!¶
Test your algorithm at least six images on your computer. Feel free to use any images you like. Use at least two human and two dog images.
Question 6: Is the output better than you expected :) ? Or worse :( ? Provide at least three possible points of improvement for your algorithm.
Answer:
In [33]:
#The output is better than I expected
In [34]:
for path in sample_files:
dog_breed_detector(path)
That's a human, but it looks like a Silky_terrier
That's a human, but it looks like a Silky_terrier
That's a dog. Breed: Bichon_frise
That's a dog. Breed: Golden_retriever
Can't detect anything!
That's a dog. Breed: Finnish_spitz
Can't detect anything!
Can't detect anything!
That's a dog. Breed: Bichon_frise